by Michael Swanwick

Back in the day, when we were both scruffy and ambitious new writers of no particular renown, William Gibson sent me a page from a novel he was writing called Neuromancer. We were long-distance collaborating on a short story at the time, and the page was meant as an illustration of I forget exactly what point of craft. But what struck me most about it was how familiar it seemed.

Jeeze, I thought. This reads just like The Man Who Melted.

Partly it was the tech. The excerpt featured a talking car pretty much identical to the cars in the novel Jack Dann was writing. Partly it was the ambience – dark, desperate, and physically degraded. But it seemed to me that Bill’s characters could have taken a detour into Jack’s novel, gambled away a few internal organs, and returned home to their own plot without the reader suspecting a thing.

The Man Who Melted didn’t make it big, the way later Dann novels such as The Memory Cathedral and The Silent would. It was much too intensely personal for that. But it’s one of those neglected books that nevertheless contain a great deal to interest the intelligent reader. At the time, I was extremely interested in it because I was witness not to its creation – that happened secretly in faraway Binghamton, New York – but to Jack’s crafting of three stories from its corpus-in-progress. Every so often, he would breeze into town with a new chunk of novel, looking for Gardner Dozois’s advice. This was before Gardner’s stint as editor of Asimov’s but everybody knew already that he was the best story doctor in the business. Together, the two of them would find a way that the section could be turned into a stand-alone story, a singleton.

And I was there.

I have perhaps written too much about those long-ago nights when God ripped the Milky Way out of the sky and flung it down at our feet while we laughed and drank bad cream sherry and plotted stories and novels and the overthrow of the reigning standards of science fiction. But I was young and those were formative times, and so I hope I may be forgiven. The significant thing is that I saw Jack craft “Amnesia,” “Going Under,” and “Blind Shemmy,” from a pre-existing work written to a different purpose.

To see how it was done, let’s compare just one story – “Going Under”– with its mother text.
In the novel, three people involved in a complex and intense relationship board the 22nd-century reconstruction of the Titanic, returning from a decadent Europe to a dying America. In mid-journey the ship will be sailed into an iceberg. Nobody knows which passengers have booked seat on the lifeboats and which have come to die. Mantle, the protagonist, is hoping to recover his memory. Pfeiffer, his unreliable friend, gives him a box containing a living mechanical head which looks and acts like Josiane, the woman he lost under traumatic circumstances and still loves. Joan, who loves Mantle, mistrusts Pfeiffer, and sleeps with both, is simply trying to hold things together. As the Titanic sinks, Joan and Mantle have a confrontation with a death-cultist who wishes to take Mantle into the darkness with her and almost kills them both. Then, at the life boats, Pfeiffer reveals that he bought a one-way ticket and runs back into the ship. Mantle tries to save him, fails, and almost dies himself.

It’s a gripping sequence of events, but to understand it you have to have read what came before. The resolution of the various issues raised is spread across the remainder of the novel.

“Going Under” opens on the Titanic with Stephen, the protagonist, meeting Esme. She is “quite young” and possibly underage. Stephen is drawn to her. She carries a box containing a living mechanical head which she calls Poppa. A precocious young boy, come to die with his elder sister, inserts himself into their affair and through his conversations with Poppa, Stephen learns that Esme intends to die. Her father’s head is programmed to keep her resolve firm should something like her newfound love for Stephen cause her to waver. When the ship goes under, he almost dies trying to save her. In the rescue dirigible afterwards, he discovers that she has survived. “Isn’t he marvelous?” she says of Poppa. “He almost had me talked into going through with it this time.”

I cannot help but wish that some small press would gather together the stories sculpted from The Man Who Melted in a single slim tome and then publish it, together with the novel, in a two-volume boxed set. There wouldn’t be much profit in it, alas. But what a great tool it would be for new writers! What a terrific opportunity to witness the creation of something marvelous, just as I was privileged to do in my youth, and to see exactly how – if you’ve got the chops – it can be done.

All three stories and the novel itself placed on the Nebula ballot for their respective years.
It was a heady experience for a writer at the very beginning of his career to watch the craft and sureness with which these works were created. It seemed to me then a wholly admirable enterprise.

It still does.

So you will understand why it grinched me to see how many reviews of The Man Who Melted referred to it as a fix-up novel.

“Fix-up” was not originally intended to be derogatory or insulting, though like so many things we say when we’re feeling lazy and grab for the handiest epithet, that’s how it usually comes out. The term was invented by A. E. Van Vogt and referred to the hack writer’s trick of taking a string of previously published stories and, with judicious rewriting and perhaps the insertion of bridging material here and there, creating from them a novel. Or, as it is most commonly described in such discussions, a “novel.”

As Samuel Johnson more pithily observed, there is nothing inherently wrong with writing for money. I am credibly informed that there are only a hundred or so individuals alive today who make a living from writing in the sf-fantasy-horror genre, without a day job, editorial work, independent wealth, or the support of Academia. This number would include both Steven King and the poor unfortunate who lives in a cardboard box and eats cat food but has at least not sunk so low as to read other people’s manuscripts for pay. And the number was significantly lower in Van Vogt’s heyday. So it would be extremely mean-spirited to begrudge even one of those writers a source of income that might be the only thing keeping him or her from grading undergraduate essays on Stanislaw Lem.

Moreover, the term, as originally intended by Van Vogt, is a useful one. It explains the loose and episodic structures of The Voyage of the Space Beagle and The Weapon Shops of Isher, which might well be critically puzzling if one didn’t know that the real money at the time Van Vogt was writing lay in the magazines, and that the income from any ensuant books was supplementary. The stories defined the ultimate shape of the novel.

This is a perfectly respectable form, whose exemplars include Clifford Simak’s City, Theodore Sturgeon’s More Than Human, and Isaac Asimov’s Foundation Trilogy.

So there is nothing wrong with calling something a fix-up, provided only that it is one.

But how often is the term properly applied? It has, sadly, undergone “critical creep,” broadening and softening like a wedge of overripe brie accidentally left out on the carpet overnight so that it has overrun its proper boundaries to the benefit of neither shag nor cheese. In a column that originally appeared in SFX magazine but can be easily googled on the Web, the scholar and wit David Langford defines “fix-up” as a process “where short fiction is shoved into irrelevant context or crudely welded together to fill out a book.” Not only does this transform a neutral descriptor into a negative value judgment, but it also defines Ray Bradbury’s The Martian Chronicles – whose components are neither irrelevant, nor crudely welded together, nor intended primarily as padding – out of a category of which it had previously been one of the crown jewels.

The critical stain runs in the other direction as well, claiming for its own books that have nothing to do with the processes Van Vogt described.

My first novel, In the Drift, was a fix-up. I make no apologies for that. I was desperate to write a novel and by dint of incredible labor managed to do so – just barely. But neither will I defend it. Like most writers, I have conflicted feelings about my first book. It appeared, it took its lumps, it went out of print, and now it is mercifully forgotten.

Nevertheless, it was my only fix-up novel.

A strict constuctionalist would deny this, pointing out that the opening segment of The Iron Dragon’s Daughter was published separately as “Cold Iron,” and that a chapter of Bones of the Earth originally appeared in significantly different form as “Scherzo with Tyrannosaur.” The first was done simply because the structure of the novel (a recurrent spiral, like Yeats’s gyres) was such that the opening movement could be published as singleton story without any alterations. So I did. On the other hand, my dinosaur stories were largely written as sketches, to help me work out the mechanics of the latter novel’s premises. “Scherzo with Tyrannosaur” proved to be such a good vehicle for conveying information that I reworked it as one of the chapters of Bones of the Earth. I moved it to a different age, located it underwater, replaced the tyrannosaur with plesiosaurs, changed the plot and several of the characters, and the conclusions the reader was meant to draw from it were completely altered, but otherwise, yes, I freely admit that it was exactly the same.

Neither of these examples comes anywhere close to meeting Van Vogt’s definition, but by the iron standards of contemporary criticism both are fix-ups. It’s like having an African ancestor – it only takes one to get you second-class service in certain restaurants. Nevertheless, I deny that the term is in either case valid.

But why stop there? I deny as well that my upcoming novel, The Dragons of Babel, despite the fact that five reworked excerpts, “King Dragon,” “The Word That Sings the Scythe,” “An Episode of Moondust,” “Lord Weary’s Empire,” and “A Small Room in Kobaldtown,” have been published separately. This I did for the money, to be sure, but also for the attention. “King Dragon” appeared in two fantasy and one science fiction best-of-year volumes, “Lord Weary’s Empire” has been picked up by another, I have reasonable hopes for “Kobaldtown,” and all the stories have received positive reviews in those rare but valued venues that cover short fiction. If the good opinion of the core genre readers is valuable (and I believe this to be self-evident), then breaking out these stories can only help the novel’s eventual future.

But the novel came first. To call it a fix-up is essentially to render a useful term meaningless.

Perhaps we need new terminology for what Jack Dann and I did with The Man Who Melted and The Dragons of Babel respectively. Which is to take sections of the book as it’s written and revise and alter them after the fact so they can be sold as independent stories. It’s as different from the fix-up process as water flowing downhill in a stream is from water being carried uphill in a bucket because the shape and substance of the stories have no effect whatsoever on the novel.

There is a time-hallowed verb for this, of course – “to cannibalize.” One cannibalizes an existing work, usually a novel-in-progress, in order to create saleable material. (The great hacks who did so much of the stoop labor involved in creating our field were past masters of this skill.) But unless you accept the awkward and unlikely “cannibalization,” there is no noun for the resulting product.

You could, if you felt the need, call the smaller works offspring fiction, since they’re the direct and legitimate children of the mother text. But there is no special term for the body of which is cannibalized, nor is there a need for one. It is simply whatever it is: a novel, a novella, an unfinished manuscript, whatever.

To call the mother text a fix-up would be like saying that a granite outcrop is made of paving stones.

In The Encyclopedia of Science Fiction John Clute anticipates, in part, my argument. It is not necessary that a fix-up be assembled from previously existing work, he states: “A book which is written so as to be broken up for prior magazine publication may well, in our view, constitute a perfectly legitimate example of the form, though we do recognize that when we call such a text a fix-up we are making a critical judgment as to the internal nature – the feel of that text.”

Which is true, so far as it goes. Publication is incidental to form, and a rose is a rose whether it finds a home in Asimov’s or not. But in theory it does not go far enough, and in practice it goes too far.

I do not wish to set up Clute as a straw man for my rhetorical Unterfussgetrampeln. His entry did not create the troublesome misapplication of a useful term. It is simply the most lucidly-stated synopsis of current norms. Its chief weakness lies in its ascribing motives to an author which cannot be documented. But its chief sin is that once the labeling of fix-ups has begun, there is no way of delimiting the results.
Let’s try a thought experiment. Here’s a story which I created specifically for this essay:

The End of All Things

On a dreary Tuesday in late November, the Time Traveler locked the room containing his machine from the inside, against the chance that somebody might rearrange the furniture in his absence and so, because two objects cannot occupy the same space, prevent his return. Then he headed for the end of the world.

He came to a time when all trace of the moon had vanished. The earth had come to rest with one face to the sun. The sky was no longer blue but black, and out of the blackness shone brightly and steadily the pale white stars. Overhead it was a deep Indian red and starless, and south-eastward it grew brighter to a glowing scarlet where, cut by the horizon, huge red-hot dome of the sun had come to obscure nearly a tenth part of the darkling heavens.

There were no breakers and no waves, for not a breath of wind was stirring. Only a slight oily swell rose and fell like a gentle breathing, and showed that the eternal sea was still moving and living.

It was so dismal that he shivered.

The red beach, save for its livid green liverworts and lichens, seemed lifeless. And now it was flecked with white. A bitter cold assailed him. Rare white flakes ever and again came eddying down. To the north-eastward, the glare of snow lay under the sable sky and he could see an undulating crest of hillocks pinkish white.

Nothing moved in earth or sky or sea. The green slime on the rocks alone testified that life was not extinct. The Time Traveler fancied he saw some black object flopping about in the distance, but it became motionless as he looked at it, and he judged that his eye had been deceived, and that the black object was merely a rock. The stars in the sky were intensely bright and seemed to twinkle very little.

The darkness grew apace; a cold wind began to blow in freshening gusts from the east, and the showering white flakes in the air increased in number. From the edge of the sea came a ripple and whisper. Beyond these lifeless sounds the world was silent. It would be hard to convey the stillness of it. As the darkness thickened, the eddying flakes grew more abundant, dancing before his eyes; and the cold of the air more intense. One by one, the white peaks of the distant hills vanished into blackness. The breeze rose to a moaning wind. All was rayless obscurity. The sky was absolutely black.

The cold, that smote to his marrow, and the pain he felt in breathing, overcame him. he shivered, and a deadly nausea seized him. He felt giddy and incapable of facing the return journey. As he stood sick and confused he saw again the moving thing upon the shoal—there was no mistake now that it was a moving thing—against the red water of the sea. It was a round thing, the size of a football perhaps, or, it may be, bigger, and tentacles trailed down from it; it seemed black against the weltering blood-red water, and it was hopping fitfully about. Then he felt he was fainting. But a terrible dread of lying helpless in that remote and awful twilight sustained him.

In that instant, the Time Traveler swore to himself that he would never again waste another precious second in the “getting and making” of life, but that he would devote himself entirely to matters of genuine consequence. In such a frame of mind, he returned to his own time. The machine faded from the all but lifeless beach. Silence again reigned supreme.

But it was particularly unfortunate that the room to which he returned was as impregnable as any fortress, and that the fuel cells upon which his now-depleted machine ran were stored without it. For on the sand, the light of the swollen sun gleamed ruddily on a metal ring – the ring which had fallen unnoticed from the Time Traveler’s pocket and which he would never again recover, holding all his keys.

*

The above was crafted from Chapter 11 of H. G. Wells’ The Time Machine. I added a paragraph and half a sentence of my own invention to the beginning, and two paragraphs to the end. In between, I cut out a great deal, and lightly reworked a couple of sentences, chiefly in order to make the story short enough to fit comfortably within this essay.

Let’s not argue about the quality of the story. I readily admit that I have carved a minor work of art from the flank of a major one. But that is the nature of a thought experiment of this sort. And here is the question it was meant to raise:

Is The Time Machine now a fix-up novel?

By current consensus, yes.

Other than his iconic stature as a founding father of science fiction, I had no particular reason for choosing Wells. I could have as easily constructed stories from the novels of Jules Verne, J.R.R. Tolkien, Norman Mailer, F. Scott Fitzgerald, Jane Austen . . . In fact, the novels that I could not pull this stunt on form a distinct minority. (I am, at a minimum, one clever sonofabitch.) Does this make The Lord of the Rings a fix-up? Or Pride and Prejudice? And, if so, does that mean that the vast majority of mimetic and fantastic novels have secretly been fix-ups all along? Because if it does not, I must somehow, via some arcane form of quantum literary entanglement, be retroactively changing the nature of their authors’ creations. Which means that I have made causality run backwards.

Which is, of course, nonsense.

And the problem with the widespread use of a critical term whose understood meaning is nonsensical is that it then serves as a barrier to thought.

Consider the episodic structure of Joe Haldeman’s The Forever War. An interstellar war, pointless as it turns out, is being fought at relativistic speeds, so that a soldier’s term of service will last subjective years but will conclude with his or her return to a society decades or even centuries different from that which was left behind – one, moreover, that has no place for the soldier, leaving little option but to re-up. Which necessarily means that for the protagonist, William Mandella, shipping out with a new (and differently equipped and trained) outfit each time, with all or most of his old friends and lovers dead or lost to the realities of relativity, every term of duty is shaped like an independent story. A situation which Haldeman took advantage of by selling four separate sections to Analog.

By anybody’s understanding save Van Vogt’s and my own, The Forever War is a fix-up.
Which makes it inconvenient that the book is nothing of the sort. Because there are so many variant explanations afloat in the noosphere regarding its origins, I wrote Joe Haldeman asking how it came about. He replied:

The first part of THE FOREVER WAR went out as a novella called “Hero” (which was also the original title of the novel; I changed it after a conversation with my brother) – but it had been conceived as a novel after the first couple of pages – which is why I always cringe at seeing it characterized as a “fix-up.” It was never broken.

I’d first sent the novella to John W. Campbell at Analog, and he wrote a really vicious rejection letter – good grief, the idea of women joining the army – and when Ben took over the editorship of the magazine and saw a carbon of the letter, he wrote or called and asked to see the story. The novel was about half written then, and he took both “Hero” and another piece of it. He eventually printed most of the book as novellas and novelettes, though he turned down one part as being too downbeat.

So I didn’t take a whole novel and chop out novelettes for Analog; Ben more or less bought the novelettes as they were written.

Let’s look at what happens when The Forever War is compared to a Haldeman novel which actually is a Van Vogtian fix-up. All My Sins Remembered was constructed from a pre-existing series of dark adventure stories about Otto McGavin, a resourceful interstellar spy. It works well enough. McGavin’s predicament (he is programmed to forget the horrific acts he performs on behalf of his government, save under hypnosis, but his mental safeguards are breaking down) grows across the arc of stories to its inevitable conclusion. Which ending, by my reading anyway, makes for a better book than had it simply been published as a collection of stories. Structurally, however, it’s loose and episodic. Stories could have been added or subtracted without significantly changing the experience.

The Forever War, by contrast, is tight as a drum. The very qualities that made it possible for Haldeman to publish so much of it in Analog, are intrinsic to the story. They intensify the reader’s involvement with the fix Mandella is in. They’re an important part of the science fiction idea. Without that structure, the novel would not work as splendidly as it does.
To think of it as a fix-up is to miss how beautifully crafted it is.

Or, considerably closer to home, consider the startling discovery I made while Googling a specific reference for this essay that an online review casually refers to Stations of the Tide as a fix-up. Now, not only were no portions of that novel published separately (nor did I try), but I swear upon my very soul that in no way was the novel written “so as to be broken up” for such publication. The thought never crossed my mind.

So what qualities in the text caused the reviewer to make such a mistake? Well . . . I think I can make a good guess, but it’s not my place to provide critical exegesis for my own work. That’s what other people are for. But – potentially – there was an interesting observation to be made there, which was neatly prevented by the misapplication of “fix-up.”

This is my objection to the current misuse of that dread term. Earlier I said that my argument was not with John Clute. And it isn’t. Whether one agrees with his latest critical judgment or not, there is no denying that he always considers the work in question for and as what it is. Which is precisely what the persistent misuse of a once-precise term prevents.

The works I’ve dealt so far fall into three formal categories: Fix-up novels, cannibalized (or “real”) novels, and mosaic novels. The term “mosaic novel” was invented by George R. R. Martin for the Wild Cards books because he wanted something to suggest that they were much more tightly interwoven than the typical shared world anthology of the time, and went well beyond merely setting a group of stories against a common background. Other works that might possibly fit this definition at the time included the Thieves World series and the Heroes in Hell books, though the creator of those latter, Janet Morris, preferred her own coinage, “braided meganovel.” More recently, authors such as Jeff Vandermeer, Richard Bowes, and Zoran Zivkovic have appropriated mosaic novel as a label for their own single-author books, meaning by it collections of stories and/or fictive artifacts that were specifically written to fit together in such a way that the whole forms a (metaphoric) picture not visible at lesser resolution.

All three forms share superficial aspects that make it easy to see why one might be confused with another. On a Venn diagram The Martian Chronicles would fall mostly within the mosaic novel circle but also partially (because, to my eye at least, it is clear that at least some of the stories were written before Bradbury came up with the idea of creating a unified book from them) in the fix-up circle. But things get considerably more interesting when one considers books commonly lumped in as fix-ups that bear few formal or intentional similarities to the above categories.

Consider, for example, Gene Wolfe’s The Fifth Head of Cerberus. It consists of three novellas which were obviously written in order to contradict, comment upon, and enlighten one another. “The Fifth Head of Cerberus” is a lush memoir of childhood and adolescence written in the first person, as by a man who is destined to be killed by his clone-son as he himself once killed his clone-father and his clone-father killed the clone-father before him. “‘A Story’ by John V. Marsh” is ostensibly written by a minor character from the first novella, but may well be evidence that the shape-shifting aboriginal species that human beings apparently exterminated when they came to the twin planets Sainte Anne and Sainte Croix have actually survived, disguised as human beings. Finally, “V.R.T.” follows a corrupt intelligence officer as he samples, almost randomly, the evidence gathered concerning a political prisoner who may or may not be John Marsh and decides that he’s probably innocent but should be executed anyway.

I am not going to attempt to pick apart the riches of the book here, other than to observe that each section illuminates and deepens the mysteries of the others in such a manner that the essential unity of the whole is inescapable. The dust jacket of the original edition of The Fifth Head of Cerberus describes its contents as “novellas” and the book as “this collection.” Yet if you visit Amazon.com and scan its reader reviews, you’ll see that almost invariably, after commenting on the structure, its readers refer to the work as a novel.

Nevertheless, by any sensible definition of a novel – even a mosaic novel – The Fifth Head of Cerberus does not belong to that category. It is neither a novel nor a collection. It is something else.

I’m going to call that something else a chimera, after both the patchwork beasts of myth and the organisms created by combining genes from two or more species. It is also – as I realized only as I was typing the previous sentence – the title of John Barth’s Chimera, itself an exemplar of the form.

Is this term necessary? The increasingly common application of “fix-up” to works that are anything but suggests that it is. When people persistently make such an error, it is an indication that they are reaching for a word which does not exist and settling for whatever lies handiest. This, happening here, is a very interesting phenomenon indeed. And it feeds upon a literary tradition that goes far back in American letters, if not necessarily in English.
The mother of this tradition may well be Gertrude Stein, whose Three Lives was finished in 1906 and first published in 1909. Austere and yet, for its famously uncompromising author at least, accessible, Three Lives is made up of three fictive biographies of working-class women in a thinly-disguised Baltimore, “The Good Anna,” “Melanctha,” and “The Gentle Lena.” Individually, the stories reveal ordinary lives to be anything but. Taken together, they say something very large about a time and a situation and a gender which were in no way particular to Baltimore. A century after it was written, Three Lives remains a work of extraordinary power. But what is important to this essay is that much of its power derives from its innovative and at first blush unshapely form. Melanctha’s life, the literal centerpiece, was given twice the length of Anna’s and three times that of Lena and, though it would have been the easiest thing in the world to intertwine their stories, the women never meet. The separation between the stories insists upon it. Their disparate lengths claim for the tales the authority of life rather than of craft. Merging the works into one well-made novel would have vitiated these effects.

There are things, the book demonstrated, that cannot be said either in stand-alone stories or in a novel. Yet they can be said.

Three Lives was an important book. Ernest Hemingway’s In Our Times was obviously influenced by it. Sherwood Anderson said it helped make him into a writer, and went on to create the single best known mosaic novel to date, Winesburg, Ohio. And over thirty years after it came out, Faulkner published Go Down, Moses which, if it was not directly influenced by Stein’s book certainly benefitted from being gestated far downstream from it, at a time when its lessons would have been understood by any serious American litterateur.

Go Down, Moses was originally published, erroneously, as Go Down, Moses and Other Stories, an editor’s mistake that Faulkner made certain would not happen again. So we know he was serious in describing the book as a novel. Nevertheless, it was an easy error to make. With seven independent stories (including “The Bear,” which was possibly Faulkner’s most frequently reprinted) written in a variety of voices and manners, it undeniably looked like a collection of short fiction.

But it was not. Just.

I do not wish to add yet one more English 201 term paper to the mountain that this book has generated. So I will simply note that Go Down, Moses is made up of seven stories; that the first (“Was”) savagely demolishes the Southern myth of plantation gentility so that the Great Mythologizer can erect a new myth of his own making on its ruins; that the second story (“The Fire and the Hearth”) is, ironically enough, itself a fix-up of three sequential stories; and that, for full appreciation, each story relies heavily on its on its predecessors. Which is to say that the book is intended to be read from front to back, in the order the stories are presented.

This last is not a minor point. Every collection of my short fiction I have ever assembled, from Gravity’s Angels through A Geography of Unknown Lands, Tales of Old Earth, Moon Dogs, and the forthcoming-as-I-write-this The Dog Said Bow-Wow – even, indeed, the five-story Croatian collection, Pet Racketa – has been carefully constructed so that the reader who starts at the front and works stolidly through to the end will start on a high note, travel through a variety of tales arranged so as to mix upbeat and downbeat, frivolous and serious, fantasy and science fiction, in a pleasing tempo, ending again on something particularly strong. Yet the readers who simply dip in and sample as they will cannot be said to be missing anything important in their experience of the book. The same goes for my collections of short-shorts, Puck Aleshire’s Abecedary, Michael Swanwick’s Field Guide to the Mesozoic Megafauna, Cigar-Box Faust and Other Miniatures, and The Periodic Table of Science Fiction. The only exception to this is The Sleep of Reason, which paired a story with each etching of Goya’s Los Caprichos. Since this last contained several recurring characters whose plot-lines evolved over several stories, entwining and sometimes intersecting with one another, the whole of which led to a final conclusion, however, it was clearly something other than a collection. (Nor, obviously, is it a fix-up, given that it was written as the installments appeared and reached its final form when the last one was posted.) One might, in a puckish mood, call it a braided mininovel, or else, because it appeared in weekly installments on the Web, a serial novella. But to treat it as merely an assemblage of stand-alone stories – as one or two time-pressed online short-fiction reviewers did – rather than reading it through from start to finish, is to do an injustice not only to the whole but to its component parts as well.

So Faulkner’s book is a novel, and perhaps it is a fix-up. If so, it is also one of the national treasures of American literature and thus, if any such thing were needed, in itself an unanswerable justification of the practice. If not, it is some related form – a matrix novel, a chimera, a whatever – and that form is essential. It frees the author from the tyranny of the protagonist and from the absurdity of the Grand Unified Plot as well. It shifts the reader’s synthetic response to the book from the particular to the universal, while grounding it always in the individual. With this tool, Faulkner was able to create not merely a myth but a world.

Whatever its taxonomy, and draw the cladogram how you will, Gertrude Stein’s cat left the bag long ago and its descendants in all their mutant forms are everywhere to be found, some as tidily well-made as Jessamyn West’s Cress Delahanty, and others as sprawling and patchwork as Ray Brandbury’s Dandelion Wine. You can call them names, but you cannot make them go away.

In light of all of the above, let’s look afresh at Thomas Disch’s 334, invariably characterized as a fix-up, as if one of the greatest stylists in the genre couldn’t have managed a more novel-like shape than this had he thought it artistically desirable. Six interwoven stories explore the lives of people leading lives of drab desperation in public housing at 334 East 11th Street in New York City in the near future. The first opens with a student enduring a lecture on Dante’s which he cannot understand, though he is in fact trapped in one of the lower levels of the hell of a morally and culturally bankrupt society. The next four explore further circles of that terrestrial Inferno inhabited by similarly blinkered denizens. Then comes the last story, “334,” which may well be formally unique. In an interview by Joseph Francavilla (Science Fiction Studies, #29, Volume 10, Part 1, March 1983), Disch explained:

My book 334 has an ambitious, conscious, formal structure. The last part of it has a three dimensional grid system which relates to and orders all the elements in the last novella of the book. And if you sit down and figure out what that grid system represents in terms of how it determines the progress of the story, and then look at the story, it does something that great formal poetry does. You don’t know that you’re being controlled by this incredible, intellectual apparatus that is totally artificial—you just read through it. And the challenge and reward of working with an artificial form is that you have to pay such attention to the smoothness of continuity of the “meter,” as it were, that the reader glides past these moments. So a difficult form simply tends to create a larger challenge, and if that challenge is met, the form vanishes before the reader. But I just couldn’t resist putting the diagram into the book anyhow. It’s meant to be a Friday’s footprint on the sand.

The 3D matrix is laid out three by three by four: bottom to top, three years (2021, 2024, 2026); back to front, three people (Mrs. Hanson, her daughter Lottie, her other daughter Shrimp); right to left, four modes of storytelling (monolog, reality, fantasy, another point-of-view). The chart tracks the three characters through one fantasy situation apiece (teevee, a museum recreation of an A&P, a nurse fetish), stays with Shrimp as seen by other eyes up from 2021 to 2024, then shows Lottie through other eyes in that same year, and so travels in an unbroken line that covers all thirty-six possible nodes in forty-three incidents (some nodes are presented twice, when the narrative line crosses itself). At the end, for reasons the reader by then well understands, its three protagonists flee into the madhouse, into religion, and into death.

Taken as a whole, 334 is not only a savage satire of the welfare state, but a clear-eyed dissection of the banality of the times. Disch’s target was not his individual characters but the society which engulfs and shapes them. By utilizing a melange of voices and situations, he created a portrait of a world made Hell.

This begins to look a lot like what Faulkner was up to. And if Faulkner is not the first writer who leaps to mind when one thinks of Disch . . . well, still, he’s no bad company to be in. The comparison invites us to take both writers a little more seriously, and to think of their works more in terms of what they were trying to accomplish than by whichever categorical closets we plan to dump them into.

If it was not a fix-up what was it? as our dear old foremother Gertrude might have said. To which I reply: a mother text with offspring fiction, a mosaic novel, a chimera, or some happy miscegenation of any or all of the above. I readily admit that by breaking a single over-inclusive compound noun into a welter of old and new terms I have confused the boundaries between categories. But is this a bad thing? When categories overlap, one is freed from the reflexive impulse to label and dismiss and given that most gracious of second chances, the opportunity to judge a work anew according to its merits.

Now that our toolbox contains more than simply a hammer, let’s try it out.

Isaac Babel’s most famous work is The Red Cavalry Stories, ostensibly nothing more than a collection of stories with a common setting and recurrent characters – the Soviet invasion of Poland in 1920 and the soldiers and civilians caught up in it. By any measure, it is a major work of literature, terrifying, moving, and a judgment on the human condition. Babel was involved in the war as a propaganda officer, and spent much of his time trying to prevent Cossacks from executing their prisoners. From the atrocities, rapes, and casual murders he witnessed, he created something of enormous depth.

Yet not all of the stories are impressive as stories. Some are vignettes or even anecdotes. They grow in cumulative power as the book is read, events recur, people show themselves in different aspects. This is an effect that relies heavily on the stories being read in the order presented. (Babel wrote more Red Cavalry stories after the book’s publication; when they are included, they are grouped separately, as afterthoughts, so as not to interrupt the original structure.) Read randomly, they would still impress and terrify. But the work as whole would be greatly diminished.

What makes this particularly interesting is that the stories themselves are seemingly presented in only the loosest of order. A story begins to tell one tale and then is interrupted and goes haring off after a totally different one. Narratives begun in one story are dropped abruptly, only to be picked up again later in the book. Events appear out of chronological order. Characters disappear and then reappear, sometimes greatly altered and other times heartbreakingly unchanged. Some never turn up again, and the reader may or may not learn what becomes of them. The narrative intelligence darts from memory to memory, never lingering long, fleeing from one to another like a sleeping man trying to dream his way out of a nightmare.

Taken as a whole, The Red Cavalry Stories looks like nothing so much as the fragments of a novel which cannot be written.

There is a scene in Federico Fellini’s Satiricon set in a workshop where Roman artists are creating fragmentary mosaics and statues without arms or heads. Babel’s book can be best understood as that same artistic project taken seriously rather than as a throwaway joke. It is the exact opposite of a fix-up. It is a novel whose continuity has been shattered by the enormities that the author witnessed.

The novel is literature’s ultimate expression of moral sense made structure, a summation and universal comprehension of the world. So when there is no sense and can be no comprehension, it is inadequate to the task and the artist needs a new form. Call The Red Cavalry Stories a mosaic novel if you wish or a chimera if you will. But it is not merely a collection of short stories.

Nor – need I add this? – is it a fix-up.

Or consider Brian Aldiss’s Cracken at Critical (Kerosina Books, 1987). Subtitled “a novel in three acts” (many of the books under consideration have similar subtitles, clearly intended to defy anticipated criticism), it lies in the region where the chimera and fix-up categories overlap. It is a chimera because it is an artistic whole made up of three stylistically disparate works with almost no effort made to disguise that fact. It is a fix-up because almost thirty years separates the first publication of the oldest story,“Equator,” and that of the newest, “Mannerheim,” and because the latter has been broken up into several pieces in order to incorporate the other stories into its corpus.

In an alternate world in which Hitler triumphed and England fell without a fight, a Finnish composer, returning home from a concert, finds a young woman’s corpse and carries it home. While waiting for the police he reads the first of two “Maybe-Myths,” as they’re called in his world, that the woman was carrying in her knapsack. During a break in his interrogation he reads a second. Both were, of course, written by Aldiss.

Here, Aldiss has treated his own works as found objects and, in much the same way that Alexander Calder might have taken a machine screw and a chunk of glass and created around them a pendant or necklace to present to one of his casual loves, carefully shaped a third work about the earlier stories, thus giving them new life and context. “‘The Impossible Smile’ by Jael Cracken” was originally serialized in two parts in Science Fantasy magazine as “The Impossible Smile” under the pseudonym of Jael Cracken. Its virtues – a genuinely creepy portrayal of a fascist Britain and some first-rate writing and observations by a first-rate mind – are swamped by a potboiler plot with hair’s-breadth escapes, corridor shootouts, lunar helicopters, and Ubermenschen telepaths. It is, in fact, as the intellectual protagonist of “Mannerheim” observes, classic consolatory fiction. And yet, that same character muses, when times are difficult, what is wrong with a little gentle consolation?

“Equator,” the second inserted story, was published as half of an Ace Double and its farrago of abductions, sudden revelations, shoot-outs, and unlikely escapes makes it even harder to defend on its own grounds than is “The Impossible Smile.” Nevertheless, when the composer – a man so high-art he expects his symphonies to be understood by only an elite few and sternly disapproves of playfulness – is confronted with his wife’s betrayal, a respected critic’s corruption, and his own nation’s treachery, he can only moan, “Put me back in my cell with my Maybe-Myths.”

The encompassing segment, “The Mannerheim Symphony,” is so beautifully and compelling written as to be a reproach to the pulp novellas embedded within it. Yet after reaching a surreal peak in which the composer’s interrogator, Captain Hakkennon, is confused with a reindeer, the plot abruptly collapses into exactly the sort of sudden revelations, reversals, and info-dumps as drove the Maybe-Myths. It ends with two single sentence paragraphs:

I was going to be an exile again.

It was my natural element.

Which are so evocative that one immediately skips back to the beginning to see what hides in the first sentence, and discovers: “The sound of dogs barking at night, when everyone else but you is asleep.”

The dogs bark and the caravan moves on – these are the words that writers use to console themselves when critics attack their work in a manner that the creators deem particularly boneheaded. That, plus the ending, plus the contrast of the Maybe-Myths against a reality that may be more convincing but offers no solace, strongly indicate that what we have here is a metafiction on the value of genre literature. Even the title, “Cracken at Critical,” proclaims that this is essentially a critical work. One which, at the end, comes down firmly on the side of the exile, the outsider, the pulp novelist.

Picasso once combined a bicycle seat and handlebars and hung them on the wall as a bull’s head. So here. Out of two old works and ingenuity, Aldiss created a chimera. It doesn’t seem to have gotten much attention, and at least some credit for that has to go to the fact that it was published at about the same time that “fix-up” gained ascendency as a means of dismissing a book without explaining why. So is it better to consider Cracken at Critical as a meretricious piece of hackwork stitched together in order to pry a few more shekels out of the pockets of innocent readers or as a serious work of art created for much the same reasons as Aldiss’s undisputed novels are?

I think the question answers itself.

Further, recognizing the multiplicity of forms has the salubrious effect of rescuing some extremely fine work (Keith Roberts’s Pavane and The Chalk Giants leap immediately to mind) from the trash-bin of critical terminology and restoring them to the upper ranks of genre fiction.
Which, surely, can be no bad thing.

So I fully intend to continue the practice of creating offspring stories from future novels in full defiance of all reckless taxonomic tagging. Not only for the reasons sketched out above, but because it is always a pleasant thing to have a new story in print. It brightens the gloom of obscurity brought on by the months- and sometimes years-long slog of writing a novel. It elevates the spirit. It makes one feel kindly disposed toward everyone in the world.\
In which spirit, I thought I’d close by sharing with you a story I recently wrote with my good friend and fellow science fiction writer (The Fixed Period), Mr. Anthony Trollope.

The Mental Daguerreotype

It was Anthony Trollope who invented the mental method of daguerreotype by which the character of men can be reduced to writing and put into grammatical language with an unerring precision of truthful description. How often previously had the novelist felt, ay, and the historian also and the biographer, that he had conceived within his mind and accurately depicted on the tablet of his brain the full character and personage of a man, and yet nevertheless, when he flew to pen and ink to perpetuate the portrait, had his words forsake, elude, disappoint, and play the deuce with him, till at the end of a dozen pages the man described had no more resemblance to the man conceived than the sign board at the corner of the street had to the Duke of Cambridge?

The great man, who was the inventor of the post box and a bit of a scribbler himself, turned his genius to correcting this dire situation, and when he was done, discovered that he had built a device that would reduce the writer’s labors to their quiddity – indeed, made of them a minor chore which could be put to rest in the interval between breakfast and teatime, leaving the rest of the day free for more serious matters.

Alas, such mechanical descriptive skill could hardly give any more satisfaction to the reader than the skill of the photographer does to the anxious mother desirous to possess an absolute duplicate of her beloved child. The likeness is indeed true; but it is a dull, dread, unfeeling, inauspicious likeness. The face is indeed there, and those looking at it will know at once whose image it is; but the owner of the face will not be proud of the resemblance.

There was, or so Mr. Trollope reflected, no royal road to learning; no short cut to the acquirement of any valuable art. Let photographers and daguerrotypers do what they will, and improve as they may with further skill on that which skill has already done, they will never achieve a portrait of the human face divine. Let biographers, novelists, and the rest of us groan as we may under the burdens which we so often feel too heavy for our shoulders; we must either bear them up like men, or own ourselves too weak for the work we have undertaken. There is no way of writing well and also of writing easily. And this will remain true, so long as men are men.

Inspired, Trollope set out to invent a mechanical man and so replace the soon to be obsolete human race with one which would properly appreciate his device’s novels.

*

Neither this nor the Wells story has appeared anywhere prior to this essay, nor have I any intention of breaking them out for separate publication. (Though as a blue-collar writer I would be foolish to turn down an offer, were one to be made.) Moreover, by demonstrating the critical absurdity of the term under question, they are necessary components of my argument. Yet by the harsh and unreasoning standards which held sway when I began this essay, but which I sincerely hope have by now been thoroughly discredited, their mere presence makes this essay itself a fix-up.

I can only plead necessity. The pulp magazines of the Gernsback and early Campbell eras had strict word limitations and there was no genre novel market to speak of at the time. So when a novel market did open up, I had no choice but to take what I had written and plump it out with extraneous and totally irrelevant material, clumsily rewritten so as to look as if it belonged, and crudely weld it all together into one unsightly and unwieldy mass. Otherwise, this essay could never have been published, and my ever-hungry wallet would be considerably thinner.

You tell me. What else could I have done?

*

Copyright 2007 by Michael Swanwick. This essay first appeared in the New York Review of Science Fiction. Permission to post “A Nettlesome Term That Has Long Outlived Its Welcome” on the Web is granted to anybody who includes this copyright notice, provided only that the essay is made available free and without charge.