“What are you offering?”

Note: This post is the fifty-fifth installment in my author’s commentary for City of Exiles, covering the epilogue. You can read the earlier installments here. 

As I’ve noted before, writing a series of novels is a little like producing a television series: the published result, as Emily Nussbaum says, is the rough draft masquerading as the final product. You want a clear narrative arc that spans multiple installments, but you also don’t want to plan too far in advance, which can lead to boredom and inflexibility. With a television show, you’re juggling multiple factors that are outside any one showrunner’s control: budgets, the availability of cast members, the responses of the audience, the perpetual threat of cancellation. For the most part, a novelist is insulated from such concerns, but you’re also trying to manage your own engagement with the material. A writer who has lost the capacity to surprise himself is unlikely to surprise the reader, which means that any extended project has to strike a balance between the knowns and the unknowns. That’s challenging enough for a single book, but over the course of a series, it feels like a real high-wire act, as the story continues to evolve in unexpected ways while always maintaining that illusion of continuity.

One possible solution, which you see in works in every medium, is to incorporate elements at an early stage that could pay off in a number of ways, depending on the shape the larger narrative ends up taking. My favorite example is from Star Trek II: The Wrath of Khan. Leonard Nimoy wanted Spock to die, and his death—unlike its hollow pastiche in Star Trek Into Darkness—was meant to be a permanent one. Fortunately, writer and director Nicholas Meyer was shrewd enough to build in an escape hatch, especially once he noticed that Nimoy seemed to be having a pretty good time on the set. It consisted of a single insert shot of Spock laying his hand on the side of McCoy’s unconscious face, with the enigmatic word: “Remember.” As Meyer explains on his commentary track, at the time, he didn’t know what the moment meant, but he figured that it was ambiguous enough to support whatever interpretation they might need to give it later on. And whether or not you find the resolution satisfying in The Search for Spock, you’ve got to admit that it was a clever way out.

The more you’re aware of the serendipitous way in which extended narratives unfold, the more often you notice such touches. Breaking Bad, for instance, feels incredibly cohesive, but it was often written on the fly: big elements of foreshadowing—like the stuffed animal floating in the swimming pool, the tube of ricin concealed behind the electrical outlet, or the huge gun that Walter buys at the beginning of the last season—were introduced before the writers knew how they would pay off. Like Spock’s “Remember,” though, they’re all pieces that could fit a range of potential developments, and when their true meaning is finally revealed, it feels inevitable. (Looking at the list of discarded endings that Vince Gilligan shared with Entertainment Weekly is a reminder of how many different ways the story could have gone.) You see the same process at work even in the composition of a single novel: a writer will sometimes introduce a detail on a hunch that it will play a role later on. But the greater challenge of series fiction, or television, is that it’s impossible to go back and revise the draft to bring everything into line.

City of Exiles is a good case in point. In the epilogue, I wanted to set up the events of the next installment without locking myself down to any one storyline, in case my sense of the narrative evolved; at the time I was writing it, I didn’t really know what Eternal Empire would be about. (In fact, I wasn’t even sure there would be a third installment, although the fact that I left a few big storylines unresolved indicates that I at least had some hopes in that direction.) What I needed, then, were a few pieces of vague information that could function in some way in a sequel. Somewhat to my surprise, this included the return of a supporting character, the lawyer Owen Dancy, whom I’d originally intended to appear just once: it occurred to me later on that it might be useful to let him hang around. When he comes to visit Ilya in prison, I didn’t know what that might mean, but it seemed like a development worth exploring. The same is true of the lock-picking tools that Ilya examines on the very last page, which I knew would come in handy. As I said yesterday, a draft can feel like a message—or an inheritance—from the past to the future. And you try to leave as much useful material as possible for the next version of you who comes along…

“This, above all else, had saved his life…”

Note: This post is the fifty-fourth installment in my author’s commentary for City of Exiles, covering Chapter 53. You can read the earlier installments here. 

I’ve noted elsewhere that I have mixed feelings about the increasing willingness among television shows to abruptly kill off their characters. On the one hand, it discourages audience complacency and raises the stakes if we feel that anyone could die at any moment; on the other, it encourages a kind of all or nothing approach to writing stories, and even a sort of laziness. Ninety percent of the time, a show can coast along on fairly conventional storytelling—as Game of Thrones sometimes does—before somebody gets beheaded or shoved in front of a subway train. But it would have been better, or at least more interesting, to create tension and suspense while those characters were sill alive. Major deaths should be honestly earned, not just a way to keep the audience awake. At least Game of Thrones knows how to milk such moments for all they’re worth; with a show like The Vampire Diaries, diminishing returns quickly set in when characters are dispatched in every other episode. It cheapens the value of life and personality, and it starts to feel questionable on both narrative and ethical levels.

Of course, I’ve been guilty of this myself, and the way certain character deaths have been incorporated into my novels testify both to how effective and to how arbitrary this kind of device can be. Ethan’s death in The Icon Thief gets a pass: it’s a striking scene that propels the last third of the story forward, and although it works in terms of momentary shock value, its repercussions continue to define the series until the final book. (The fact that it was a late addition to the story—indeed, it was one of the last things I wrote—hasn’t kept it from feeling inevitable now.) The corresponding scene in City of Exiles, which echoes its predecessor in a lot of ways, is a little harder to defend. It’s a nice, tricky chapter, and I’m still proud of the reversal it pulls, but it feels a bit more like a gimmick, especially because its consequences don’t fully play out until the following novel. From a structural point of view, it works, and it provides a necessary jolt of energy to the story at the right place, but it’s not that far removed from the way a show like 24 will throw in a surprise betrayal when the audience’s attention starts to wander.

Looking back, I have a feeling that my own uneasiness over this moment—as well as the high body count of the novel as a whole—may have led me to spare another character’s life. Toward the end of the process, there was a lot of talk about whether I should kill off Powell. After reading the first draft, my agent was openly in favor of it, and it’s true that things don’t look particularly good for Powell at this point: realistically speaking, it’s hard to imagine that anyone on that airplane could have survived. Much earlier, I’d even toyed with the idea of killing Powell at the end of Part I, which would have made Wolfe’s journey all the more urgent. Between these two possibilities, the latter seemed much more preferable. A death at the conclusion of the novel wouldn’t have advanced the narrative in any particular fashion; we’re only a few pages from the end anyway, and if the stakes aren’t clear by now, there’s no point in trying to heighten them in retrospect. Killing him earlier would have served a clearer dramatic purpose, but I also would have lost his far more wrenching scene on the plane, which I don’t think would have been nearly as strong without him at its center.

In the end, I let him live, though badly hurt, for a number of different reasons. At the time, I thought that I wanted to preserve the duo of Powell and Wolfe for a potential third novel, although as it turned out, they don’t spend a lot of time together in Eternal Empire, and his role could conceivably have been filled by somebody else. Powell also benefited from my impulse to pull back on the death toll of plane crash: I didn’t want to kill off Chigorin, mostly because he was transparently based on a real person whose imagined demise I didn’t much feel like exploiting, so most of the other passengers ended up being protected by his character shield. Most of all, I thought that keeping Powell alive would restore a necessary degree of balance to the ending. City of Exiles concludes on something of a down note: Ilya is still in prison, Karvonen’s handler is still at large, and Wolfe still doesn’t know—although the reader does—that the traitor in her organization is someone close to her own heart. Killing off Powell would have left the situation feeling even more hopeless, so I spared him. If this all sounds a little cold and calculated, well, maybe it was. Powell might not have made it, but he escaped thanks to luck, impersonal considerations, and a moment of mercy from the universe. And that’s true of all of us at times…

“She had entered an underground labyrinth…”

Note: This post is the fifty-third installment in my author’s commentary for City of Exiles, covering Chapter 52. You can read the earlier installments here. 

The problem with writing the high points of any novel is that they’re always going to seem vaguely familiar. Authors have been working out solutions to the same handful of scenes—the chase, the game of cat and mouse, the final showdown—for centuries, and even in otherwise forgettable works, these are the moments we tend to remember. As a result, our heads are populated with images and tropes from countless previous thrillers, and after a while, they feel as if they’re ringing variations on the same themes. (This is why I find myself tuning out more and more during movie action sequences, especially the kind that rely on digital effects, as in the last act of Captain America: The Winter Soldier, a movie I loved for its first hour before finding myself increasingly detached.) Every now and then, we’re presented with a set piece that gives us something we genuinely haven’t seen before: by now, it’s a cliché in its own right, but I still remember being exhilarated by the ending of the Coen Brothers’ Blood Simple, with the villain in one room, the target in the other, and the shafts of light pouring through the bullet holes in the wall. And aside from that rare kind of inspiration, writers are left to express their personalities with little touches in big scenes.

In my own work, I’ve tried to make each action scene as distinctive as possible while still moving fluently within the beats of the genre. There’s a kind of pleasure in seeing a writer deploy familiar elements in an expert fashion, and I’m generally pleased by my efforts in that direction. Many rely on a single large idea or setting to force the action into a more unusual shape. The climax of The Icon Thief, in which Maddy breaks into the Étant Donnés installation at the Philadelphia Museum of Art while being unknowingly stalked by a killer, still strikes me as a neat idea, and most of the story was designed to bring her, and us, to that exact moment. In Eternal Empire, which I feel has the best action writing of the entire series, the logic of each turning point was determined by a distinctive setting: the London riots, a wedding, a sinking yacht. And you can see a progression there, as I had to widen the net ever further to make the material seem fresh. After only three novels, it felt increasingly hard to stage scenes like this in ways I hadn’t done already, which is part of the reason I’ve tried to move in other directions in my writing. Stick long enough to any one mode, and you inevitably start to repeat yourself.

City of Exiles feels like a transitional novel in more ways than one. The action here is still relatively grounded: if some of the big sequences in Eternal Empire have a cast of hundreds, this story keeps it intimate and intense, even when the stakes are enormous. (Subconsciously, I may have also had the arc of the overall series in mind: I confined the most wrenching moment in the story to a private plane because I knew I had a megayacht on the horizon.) It was also my first chance to really play with the conventions of a certain kind of crime procedural. Law enforcement officers like Wolfe and Powell occupy an important supporting role in The Icon Thief, but that story was primarily about an ordinary woman who couldn’t be expected to carry a shootout or car chase. City of Exiles is in some respects a more conventional novel—although still undeniably peculiar—and it allowed me to indulge in correspondingly straightforward action. Much of the novel reflects the experiences of a writer who has suddenly been given a new set of toys, and I relished the chance to write about SWAT teams, surveillance, and the exchange of fire in close quarters between two antagonists who are equally armed, proficient, and desperate.

That’s particularly true of the climax, which occupies Chapter 57, the longest single chapter of the novel. Even if the reader isn’t clued in by the rhythms of the story itself or the dwindling number of pages, there are plenty of structural signs that we’re nearing the end: unlike most chapters, which stick to a single character’s perspective, it switches three times between points of view, moving from Karvonen to Wolfe and back again. I try to save this kind of crosscutting for extended action scenes that couldn’t be divided up without sacrificing momentum, and I’d like to think that the reader picks up on this—the scene starts fast and keeps going, as if we’re rushing headlong to a decisive moment. Since this is a novel with multiple echoes, intentional and otherwise, of The Silence of the Lambs, perhaps it’s inevitable that it would end in a similar way, and the chapter’s final page, in particular, owes a lot to Harris. But I still love the result, even now, when my feelings as the novel as a whole are still evolving. There are times when I think City of Exiles it the best novel in the series, and others when I suspect it may be the weakest. Certainly its seams show a little more than in its predecessor or successor. But whenever I think of it as my best, it’s because of scenes like this, with the familiar and the unexpected colliding in one last confrontation in the dark…

“This is where he wanted to go…”

Note: This post is the fifty-second installment in my author’s commentary for City of Exiles, covering Chapter 51. You can read the earlier installments here. 

Years ago, I served as an alternate juror in a civil case in Brooklyn. The details of the lawsuit don’t really matter—it involved a patient alleging malpractice, ultimately without success, after undergoing cataract surgery—and I didn’t even get to stick around long enough to render a verdict. I took good notes, though, on the assumption that the experience might be useful for a story one day. This hasn’t happened yet, but one detail still sticks with me. Part of the case hinged on what the doctor had written in the patient’s file, so at strategic moments in the proceedings, the lawyer for the plaintiff would put an enormous reproduction of the relevant page on an easel, inviting us to look closely at some marginal note in an illegible doctor’s scrawl. And what struck me was the fact that records like this are kept for every patient, filling cabinets and boxes in every doctor’s office in the country. Most end up filed away forever. But every now and then, a trial or insurance settlement will depend on detail from a past case, so one dusty file will be promoted out of storage and blown up to huge proportions. It’s a kind of apotheosis, the moment when an ordinary document turns into a key piece of evidence, and we’re asked to study it as closely as a sacred text.

You see the same phenomenon whenever a mass of information promises to yield one small, crucial clue. Conspiracy theorists pore over every scrap of paper connected to events like the Kennedy assassination, until what might otherwise be a routine report or standard form acquires a sinister significance. And writers—who differ from conspiracy theorists mostly in the fact that they’re aware that what they’re doing is fictional—often find themselves up to the knees in a similar process. When you’re writing a novel that requires any amount of research, you find yourself collecting whole shelves of material, but in practice, you find that a critical plot point hinges on a little morsel that you gathered without understanding its full importance. You’ll be trying to map out a scene, for instance, and realize that it has to take place in a particular corner of a building that you’ve never seen before, or that you visited months ago and have mostly forgotten. When that happens, you go back over your notes and sketches, look up photographs, stare at maps, and hope to find the tiny bit of data you need, which often turns on a few blurry pictures that you can barely see.

I often found myself staring at images like this. When I was writing The Icon Thief, for instance, I knew that the action of the last chapter depended on a detailed knowledge of the interior of Étant Donnés, the enigmatic work by Marcel Duchamp that was installed at the Philadelphia Museum of Art after his death. Since I couldn’t easily get inside that room myself, I was forced to depend on the sources I had, shelling out ninety dollars for a copy of Duchamp’s Manual of Instructions and going over the illustrations until I had a pretty good idea of what my character would find. (Just before the novel was complete, Michael R. Taylor published his definitive study of Étant Donnés, which had much better pictures. It was too late for it to influence the story itself, but it allowed me to correct a number of small errors.) Similarly, in City of Exiles, my description of the London Chess Classic was based on a trove of pictures from the tournament’s official website, which I used to clarify my descriptions, the layout of the building, and the logic of the ensuing chase scene. And I don’t think the photographer in question ever imagined that those images would be used for that purpose.

Ideally, of course, we’d be able to verify everything firsthand, and I’ve tried to do my own location research whenever possible. Yet there’s also something to be said for the experience of looking at a scene through a very narrow window. You can’t range freely through the world; the maneuvers you make are constrained by the evidence you have at hand, which forces you to focus and scrutinize every detail for possible use or meaning. I knew, for example, that the ending of City of Exiles would take place in the network of tunnels under Helsinki, which was something I couldn’t easily visit. All I had, in the end, were a handful of pictures and a video that offered a few tantalizing glimpses of the interior, amounting to no more than a few seconds. From those fragments, I was able to build the sequence that starts here, in Chapter 51, as Wolfe arrives at the data center that provides an access point to the tunnels. Making it plausible involved going through the footage I had inch by inch, pausing it repeatedly to figure out the geography and how to describe what I was actually seeing. Mistakes undoubtedly crept in, and I’m sure I would have benefited from walking those tunnels myself. But as it stood, I had no choice but to put together the pieces I had, put my characters inside, and see what happened when they met…

“He took in his surroundings…”

Note: This post is the fifty-first installment in my author’s commentary for City of Exiles, covering Chapter 50. You can read the earlier installments here. 

One of the few really useful tricks I’ve picked up as a writer is that if you don’t know what happens in a particular scene, try giving it a location. There’s a book on the movies—I think it’s Frank Capra’s The Name Above the Title, but it could also be Walter Kerr’s The Silent Clowns—that describes a comedian walking onto a standing set and immediately coming up with bits of business involving the furniture and props on hand, and a similar process seems to operate in fiction. When you’re inventing a sequence from scratch, whether it’s a chase scene or a quiet interaction between two characters, you’re initially handicapped because the setting in which it occurs is a blank stage. If you can assign it a location, even a relatively arbitrary one, the layout of the surroundings quickly suggests ideas for movement, action, and rhythm, or what a stage director would call blocking. And although a novelist can design a fictional location in any way he likes, in practice, it’s best if the place involved is a real one with concrete physical constraints.

This is part of the reason why so many authors enjoy drawing maps. In fantasy fiction, a map of the territory often precedes the writing of the story itself, both because worldbuilding is a fun pursuit—even without a narrative to support it—and because the landmarks can impose their own kind of logic. (There’s an entire book, Maps of the Imagination by Peter Turchi, devoted to teasing out the parallels between cartographic and narrative thinking, and it’s worth a read.) Robert Louis Stevenson went so far as to recommend mapmaking to writers of all kinds:

But it is my contention—my superstition, if you like—that who is faithful to his map, and consults it, and draws from it his inspiration, daily and hourly, gains positive support, and not mere negative immunity from accident. The tale has a root there; it grows in that soil; it has a spine of its own behind the words. Better if the country be real, and he has walked every foot of it and knows every milestone. But even with imaginary places, he will do well in the beginning to provide a map; as he studies it, relations will appear that he had not thought upon; he will discover obvious, though unsuspected, shortcuts and footprints for his messengers.

The value of maps may be less obvious for a novel like City of Exiles, but in practice, they turned out to be absolutely crucial. Much of suspense fiction, as I’ve noted before, consists of laying down an intensely detailed stratum of “realism” that allows the writer to get away with greater imaginative leaps, and that was especially the case here: the plot hinges on a series of implausible events that work only if they’ve been grounded in what seems like some version of the real world. Location research played an important role here, and the trip to London I took paid dividends in such scenes as Karvonen’s first hit and the chase at the London Chess Classic. These are scenes in which real locations dictated much the action, and I don’t think I could have invented anything nearly as convincing if I hadn’t, as Stevenson says, walked every foot and learned every milestone. And even when I wasn’t able to check out a location firsthand, I still relied on maps and landmarks, arguably to an even greater extent, since it meant that I had to plot out complicated action from an armchair.

In Chapter 50, for example, the logic of the story hinged on a solution to a specific series of geographical problems. Karvonen is driving through a snowstorm in Helsinki, heading for the passenger harbor, when he’s forced to make a detour because of a traffic accident. Along the way, he’s stopped by a police van, and in order to avoid being arrested, he shoots and kills the officer. The crime has to be witnessed, forcing him to abandon his car, but he still has to be able to slip away and head for the next place in his itinerary, the network of tunnels under the city that I knew from the beginning would be the setting for my climax. After poring over Google Maps for most of an afternoon, I finally ended up with a location that worked, near the park by Uspenski Cathedral. (Among other things, it allowed me to conveniently interpose a canal between Karvonen and the onlookers to the shooting, who could witness it without being able to respond in time.) If you read the chapter carefully, you’ll see that every beat was suggested or determined by the geography I had to follow. The result is one of my favorite scenes in the novel. And it wouldn’t have worked at all if I hadn’t had a map…

“A big, friendly officer…”

Note: This post is the fiftieth installment in my author’s commentary for City of Exiles, covering Chapter 49. You can read the earlier installments here. 

“It is a time-proven rule of the novelist’s craft,” John Fowles writes in The French Lieutenant’s Woman, “never to introduce but very minor new characters at the end of a book.” Fowles is being a little facetious here: the character whose first appearance these lines introduce is either God himself or a veiled surrogate for the author. But he makes a decent point. In what we think of as a properly constructed novel, the ending is a kind of recapitulation or culmination of all that came before, with a recurrence of characters, images, or themes that John Gardner has visualized in a famous image from The Art of Fiction. Introducing anything new at this point can feel like poor planning, and that’s especially true of the human players. Character, by definition, is revealed by action in time, and when you only have a handful of pages left to wind up the story, anyone who shows up at the last minute usually won’t have room to develop anything like a real personality. He or she feels like what the other characters might well be, but have had more of a chance to hide: a plot point, or a puppet.

Which only suggests that the rule against introducing new characters late in the story is just a particular case of a more general principle. A novel is a machine constructed to hold the reader’s attention, but the best novels keep their internal workings well out of sight. Among other things, this often involves concealing the real reason a character has been included in the story. Even in literary fiction, most characters are there for a specific purpose: to advance the plot, to illustrate a theme, to provide the protagonist with an important interaction or a moment of contrast. Sometimes a character will be introduced on page five for the sake of a scene two hundred pages later, and it’s the intervening space that makes it seem natural. When the gap between a character’s initial appearance and his or her reason for being there is reduced, we start to see the wheels turning, and that’s especially true near the climax of the novel, when the range of possibilities the story can cover is necessarily constrained. If a major character shows up fifty pages from the end, it often isn’t hard to figure out why.

What’s funny, of course, is that what seems like a departure from reality is actually a departure from a different kind of artifice. In real life, people don’t appear on schedule: enormous presences in our lives can be introduced at any time, and the sequence of events doesn’t fall into a neat pattern. We see this clearly in books or movies based on real incidents: a movie like Zero Dark Thirty struggles—very successfully, I might add—with the fact that the players in its climax are a bunch of guys we haven’t met before. It’s easier to accept this when the narrative presents itself as a true story, and a plot invented from scratch wouldn’t be likely to take the same approach. You might even say that a story that wanted to come off as factual could introduce new characters at any point, as they appear in life, but in practice, the result seems paradoxically less convincing. (This may be why when a major character is introduced late in the game, it’s often because he’s compelling enough to overcome any objection. My favorite example is Jean Reno in La Femme Nikita, who makes such an impression in the last thirty minutes that he ultimately got what amounted to a spinoff of his own.)

In Chapter 49 of City of Exiles—on page 342 of 396—I introduce a character named Timo Lindegren, a senior constable in the Helsinki police department. Shrewd readers, noticing how few pages remain in the novel and that a big chase sequence seems to be impending, might conclude that Lindegren has appeared on the scene just so he can be shot to death forty pages later, and in fact, they’d be right. I won’t pretend that Lindegren is anything other than a functional character, there to give Wolfe someone to talk to as she tracks her killer in the endgame and to die at the moment when the danger seems greatest. (He’s also there, and not trivially, to give Wolfe a handgun when she needs it.) And what strikes me now, reading these chapters over again, is that in a different version of the same novel, Lindegren might well have been introduced three hundred pages earlier, only for the sake of filling the exact same role he does here. If that had been the case, his function might not have been so obvious, but the late change of scene to Finland meant that he could only show up just in time to be knocked off. That’s essentially true of many other characters, but it feels particularly blatant here because we’re so close to the end. But he’ll stick around for a little while longer, at least before his abrupt exit…

“He knew that he had waited too long…”

Note: This post is the forty-ninth installment in my author’s commentary for City of Exiles, covering Chapter 48. You can read the earlier installments here. 

One of the trickiest aspects of writing a story, especially when you’re working within a familiar genre, is the management of nuance. Nuance, in general, is a good thing in fiction: life itself is nothing if not composed of ambiguities, and we tend to judge authors by how well they reproduce those subtleties and unknowns. Yet clarity also counts, and much of the revision process is spent trying to trike a balance between an accurate representation of the world’s uncertainties and the clean line of narrative that a readable novel demands. Nuance, in itself, can amount to a stylistic device: its homely details and smoothly rendered contradictions become a way of concealing how schematic the story really is. A mystery novel, for instance, is a sort of confidence game, an intricately designed puzzle that pretends to be an organic sequence of events. The dead ends and red herrings that the author builds into the story are as calculated as anything else, and the result only works if the sleight of hand remains invisible.

This can also apply to character, in even more insidious ways. Fiction rests on its ability to create the illusion that names on a page are real men and women, even as they occupy roles within the overall picture. Too much emphasis on one side of the equation can throw off the entire story, so writers find ways of sustaining fiction’s simulation of reality while simultaneously advancing the plot. This is why fiction places so much weight on motivation, which can be a fiction in itself. One of the few points on which most professional writers can agree is the importance of a clear sequence of objectives: as both David Mamet and Kurt Vonnegut have said, at any given point in the story, it should be fairly obvious what the protagonist wants and how he or she intends to get it. In real life, we don’t always know why we do things, and while some writers have devoted their careers to evoking that kind of ambivalence, in practice, fiction demands a little more clarity than strict psychological accuracy would allow. And much of the challenge of creating compelling characters lies in figuring out how much nuance is enough.

In many cases, it’s the story itself that provides essential clues. We’re usually told that characters should shape the plot, rather than the other way around, but in fact, it’s both permissible and desirable to have the line of influence run both ways. None of us exist in isolation from the world around us; our personalities aren’t cleanly demarcated, but blur into our interactions with others and the situations in which we find ourselves. A decision that might seem perfectly logical and considered at the time later turns out to have been shaped by outside forces of which we’re only dimly aware, and it’s only in retrospect that we start to see how we were part of a larger pattern. The ongoing dialogue between character and story reproduces this in miniature. Character only has meaning in the fabric of the narrative within which it’s embedded, and the needs of the plot can provide crucial, and surprisingly nuanced, information about behavior—often in ways that would never occur to the author if he were creating a character without any context.

In City of Exiles, for example, Lasse Karvonen is as close to a classic villain as any I’ve written. He’s a sociopath with all of the usual warning signs—including pyromania and cruelty to animals as a child—and most of his actions arise from a cold, nearly inhuman pragmatism. Yet he has one big weakness: his sentimentality toward his home country of Finland. At this point, it’s been long enough since I wrote the novel that I don’t really remember if I introduced this trait with an eye to how it would pay off later on, but toward the end of the story, it creates his first real internal conflict, as he struggles over whether to obey an order from above by eliminating his Finnish accomplice. For the most part, it seemed best to render Karvonen in shades of black and white; this is such a complicated story that it needed a storybook villain to drive the action. But I found that giving him a moment of hesitation in Chapter 48, as he decides whether or not to kill the young woman who has given him her trust, paid off on multiple levels, which is always a sign that you’re on the right track. And perhaps it’s not surprising that such a late change of heart ultimately leads to his downfall…

“Whose locker is this?”

Note: This post is the forty-eighth installment in my author’s commentary for City of Exiles, covering Chapter 47. You can read the earlier installments here. 

There’s a point in the audio commentary for one of the Bourne movies—I think it’s The Bourne Ultimatum—when director Paul Greengrass admits that he made things a little too easy. Bourne has narrowly avoided being assassinated at London’s Waterloo railway station, escaping with nothing but a dead reporter’s notebook, and he has no way of knowing who ordered the hit. Fortunately, the notebook happens to contain the name of an investment advisory firm that bankrolled the operation in question, so Bourne does what any of us would do in that situation: he googles it. He comes up with an address in Madrid, confirms it against a receipt in the reporter’s notes, and then he’s off to another big action scene. Needless to say, this all seems a bit too simple, and if we weren’t caught up in the movie, we might object to it. But Greengrass argues, and with good reason, that in this kind of story, it’s more important to move from one beat to the next as quickly and economically as possible, rather than derailing the momentum with a more plausible sequence of events.

I think he’s right. It’s easy to make fun of certain stories, especially thrillers and action movies, for the leaps of logic that the hero has to make to get from one stunt sequence to another. Even superficially more realistic procedurals are grounded less on real crime scene technique than on sudden flashes of insight, and if you were to cut all of them together, they would start to seem even more ridiculous. Yet it’s a convention that arises less out of a lack of concern about “realism” than from the set of rules that the movie itself has established. Plenty of films, from All the President’s Men to Zodiac, have made riveting cinema out of the tedium of ordinary reporting or investigative work, but they’ve been conceived before the fact in a way that prepares us for the kind of story we’re about to watch. A Bourne movie presents us with very different expectations: the only logic that matters is that of restless movement, and to the extent that the film presents certain elements more or less plausibly, it’s only to facilitate our larger suspension of disbelief. Bourne googles his way over a bump in the script because it was the most efficient way to get from point A to point B.

We see this kind of compression and elision even at the highest levels of literature. I’ve always loved what John Gardner had to say about Hamlet, which includes a moment of high implausibility: the fact that the normally indecisive prince has no trouble sending Rosencrantz and Guidenstern to their deaths offstage, and with almost no explanation. “If pressed,” Gardner writes, “Shakespeare might say that he expects us to recognize that the fox outfoxed is an old motif in literature—he could make up the tiresome details if he had to.” He continues:

But the explanation I’ve put in Shakespeare’s mouth is probably not the true one. The truth is very likely that almost without bothering to think it out, Shakespeare saw by a flash of intuition that the whole question was unimportant, off the point; and so like Mozart, the white shark of music, he snapped straight to the heart of the matter…Shakespeare’s instinct told him, “Get back to the business between Hamlet and Claudius,” and, sudden as lightning, he was back.

In other words, it’s a question, like so much else in art, of prioritizing what is truly important. And sometimes realism or plausibility takes a back seat to advancing the overall narrative.

Many of the same factors come into play in Part III in City of Exiles. The previous section ends with Wolfe in London, helpless to prevent the crash of Chigorin’s plane; Part III concludes with her final confrontation with Karvonen in a tunnel beneath Helsinki. To get from one point to the next involves covering an enormous geographical distance and an even more tenuous chain of associations. Wolfe needs to figure out that the plane was sabotaged in Finland, find Karvonen’s contact at the airport, track her down, interrogate her, and preemptively think ahead throughout to anticipate where Karvonen will go now, all in exactly fifty pages. Pulling this off in a way that also kept the story going involved a fair number of shortcuts, as we see in Chapter 47, in which Wolfe identifies Karvonen’s accomplice thanks to the lucky glimpse of a volume of John Donne’s poetry in her locker. If this feels like something of a cheat, well, maybe it is. Still, I had little choice if I wanted to keep things moving. Playing this kind of card too often can strain plausibility to the breaking point, which hurts the story more than it helps. But here, it seemed more important to get Wolfe as soon as possible to her appointment under the city…

“It’s been too long since we truly spoke…”

Note: This post is the forty-seventh installment in my author’s commentary for City of Exiles, covering Chapter 46. You can read the earlier installments here. 

There’s a moment in Bong Joon-ho’s Snowpiercer—which is one of the more interesting movies of this or any year—that demonstrates how shrewdly the film handles its own structure, even within a narrative that moves within wild extremes of realism and tone. Toward the end of the movie, the small band of rebels has fought its way to the front of the titular train, in search of its mysterious designer and engineer. We reach a point where only a single door stands between the hero, played by Chris Evans, and the engineer himself. And it’s here, with the climax seemingly around the corner, that the movie pauses. Evans sits down, lights the obligatory cigarette, and for the first time in the film, he really talks. So far, he’s spoken in short bursts of exhortation or exposition, usually just for a sentence or two, but now he opens up and finally reveals where he comes from and what brought him to this moment. I’m not a fan of backstory; I’m in full agreement with William Goldman that a hero is more compelling the less he reveals. But if you’re going to do this kind of thing, I’d argue that it belongs here, after the protagonist has been clarified through an entire story’s worth of action, but before he faces his final challenge.

In other words, Snowpiercer, a movie which otherwise moves with relentless momentum, is smart enough, even within its tonal chaos, to identify and utilize an organic pause in its structure. It’s particularly effective here, near the end: when the door opens and Evans confronts what awaits him on the other side, his personal stakes are fresh in the viewer’s mind. We’re also more likely to accept a break here than anywhere else. By now, we’re impatient to see what lies behind the door, but our awareness of exactly where the movie is going gives his monologue an urgency that it wouldn’t have if the next steps weren’t so clear. And while this may seem like a trivial point, it’s one of a very few valuable tricks—which I’ve noted elsewhere can be counted on one hand—that work in a wide range of stories. One of the recurring headaches of a writer’s life is finding room for moments where the narrative can pause and regroup. Throughout most of a novel or film, especially in the suspense or action genres, that kind of break is lethal to the forward movement that the audience expects. And I’ve found that this kind of interstitial scene, between a climactic setup and payoff, works better than any alternative.

If I seem to be giving this point more emphasis than it really deserves, it’s only because the problem of the narrative pause is one that I’ve confronted again and again, and this is one approach that really works. It may even be more effective in a novel than on film. A movie unfolds at the same number of frames per second no matter the content of the scene, so a long break just before the climax can subtly test the viewer’s patience. In a novel, by contrast, much of a writer’s time is spent managing the pace at which the reader turns pages. For most of us, reading speed is flexible; we read certain passages more slowly to absorb what they say, while the pages fly by at moments of high suspense or excitement. And this kind of momentum can carry over from one scene to the next, so that a quiet chapter that might have broken the rhythm earlier in the novel can be swept up in the reader’s eagerness to see what happens next. As usual, it takes repeated readings and revisions by the author to get the balance just right. But when it works, it propels the reader past moments of necessary consolidation that the story might otherwise be unable to accommodate.

For my last few novels, I’ve used moments like this, particularly at the beginning of a new section when a cliffhanger from the previous scene remains unresolved, to incorporate flashbacks, which I otherwise try to avoid. I began doing this on a conscious level only recently, but it appears in a germinal form in Chapter 46 of City of Exiles. By the time we reach the first chapter of Part III, we’ve just witnessed the crash of Chigorin’s plane, and the stakes for the last hundred pages are as clear as they come. Instead of picking up that thread right away, though, I pause for a relatively quiet exchange between Ilya and Vasylenko, as the two men really talk for the first time in the novel. I did this mostly for pragmatic reasons: I needed Wolfe to fly out to Helsinki to investigate the aftermath of the crash, and putting a scene here inserts a brief delay that serves as the psychological equivalent of her plane ride. But it also gave me a chance to flesh out the themes and emotions that will pay off very soon. At the time, I don’t think I was aware of what I was doing; as with most writing rules, you hit on it intuitively, then go back to figure out what you’ve done. But it was no accident that the scene takes place here, when so much else is already in motion. As Vasylenko says: “There is nothing that anyone can do to prevent what is happening already…”

“The greatest and most terrible sight…”

Note: This post is the forty-sixth installment in my author’s commentary for City of Exiles, covering Chapter 45. You can read the earlier installments here. 

One of the dangers of writing any kind of fiction, literary or mainstream, is how quickly the story can start to exist within a closed circle of assumptions. The rules of a genre aren’t a bad thing: as I’ve noted elsewhere, they’re essentially a collection of best practices, tricks and techniques that have accumulated over time through the efforts of countless writers. A trick that survives is one that has repeatedly proven itself, and much of the pleasure of reading comes from watching as the author honors, subverts, or pushes against the constraints that the narrative imposes. The trouble is when a story moves so far from the real world that its characters cease to exhibit recognizable human behavior, as its internal rules become ever more strict and artificial. A show like The Vampire Diaries, for instance, takes a surprisingly casual approach to murder, with the average episode boasting a body count in the high single digits, and the reaction to each additional death amounts to a shrug and a search for a shovel. Within the confines of the show, it works, but the second we start to measure it against any kind of reality, it comes precariously close to collapsing.

That’s true of literary fiction as well. Even great authors operate within limits when it comes to the kinds of situations and characters they can comfortably depict. In Genius and Lust, Norman Mailer draws a memorable comparison between the tonal ranges of Henry James, Ernest Hemingway, and Henry Miller:

The cruelest criticism ever delivered of Henry James is that he had a style so hermetic his pen would have been paralyzed if one of his characters had ever entered a town house, removed his hat, and found crap on his head (a matter, parenthetically, of small moment to Tolstoy let us say, or Dostoyevsky or Stendhal). Hemingway would have been bothered more than he liked. Miller would have loved it.

The more closely we read certain writers or genres, the more we see how much they stick to their particular circles. Sometimes that circle is determined by what the author can talk about through firsthand experience; sometimes it’s the result of a genre enforcing an unstated decorum, a set of rules about what can and can’t be said.

When it comes to suspense and mystery fiction, these rules can lead to a suspension of emotion, at least of certain kinds. A murder mystery never shows much regret over the fate of the departed; it’s too busy moving on to a trail of clues to waste any time in mourning. Suspense works along similar lines. Sometimes a pivotal death will serve to motivate an ensuing course of action, but along the way, the bodies tend to pile up without much in the way of consequence. I wouldn’t say that my own novels take this as far as The Vampire Diaries, but when I look back on The Icon Thief and its sequels, there are times when I get a little uneasy with the way in which the plot advances on moments of casual violence. (On a much higher level, you can hear some of the same ambivalence in Francis Coppola’s voice when he talks about The Godfather, and by the time he gets to The Godfather Part III, he seems outright weary at having to supply the hits and kills that the audience has come to expect.) There’s a mechanical pleasure to be had in seeing a story run fluently through those conventions, but when you step briefly outside, you start to see how limited a picture of the world it really presents.

That’s why I’m particularly proud of Chapter 45 of City of Exiles. It’s a short chapter, as short, in fact, as I could make it, and my agent even suggested that it be cut. I’m glad I kept it, though, because it represents one of the few points in the entire series when we pull away from the primary characters and depict an event from an outside perspective. In it, I introduce a character named Ivan, fishing on the ice with his dog, who happens to witness the crash of Chigorin’s private plane. In some ways, my decision to cut away here was a pragmatic one: none of the passengers is in any condition to directly experience what happens, and there’s a world of difference, in any case, between describing a plane crash from the inside and showing how it appears on the ground. On a more subtle level, I wanted to depart from the closed circle of the novel to reinforce the horror of the moment, even if it’s described as clinically as everything else. Objectively speaking, City of Exiles is a violent book, and there are times when the faces of the victims start to blur together. Here, for once, I wanted to suggest how it would feel to a man who didn’t know he was part of the story. Ivan won’t be coming back again, but it was important, if only for a moment, to see through his eyes…

“A freezing horror took hold of him…”

Note: This post is the forty-fifth installment in my author’s commentary for City of Exiles, covering Chapter 44. You can read the earlier installments here. 

I’ve always been fascinated by horror fiction, but I’ve rarely drawn on its conventions for my own work. A few of my short stories—notably “The Boneless One” and “Cryptids”—employ horror tropes, and “Kawataro” is essentially an extended homage to the genre. In my novels, though, there’s little if any trace of it. Part of this is due to the fact that I’ve ended up working in a category that doesn’t accommodate itself easily to that style: suspense fiction, at least of the international kind that I write, operates within a narrow tonal range, with heightened events and purposeful violence described with clinical precision. This air of constraint is both the genre’s limitation and its greatest strength, but it also means that horror sits within it uncomfortably. At its best, horror fiction comes down to variations of tone, with everyday mundanity disrupted by unknown terrors, and a writer like Stephen King is so good at conveying the ordinary that the horror itself can seem less interesting by comparison. (Writers in whom the tone is steeped in dread from the beginning have trouble playing these changes: I love H.P. Lovecraft, for instance, but I can’t say that he scares me.)

The big exception is Chapter 44 of City of Exiles, in which horror comes to the forefront of the narrative to a degree that doesn’t have a parallel in the rest of the series. City of Exiles isn’t a perfect novel, and I’ve been hard on it elsewhere in this commentary, but I still think that the last ten chapters or so represent some of the strongest writing I’ve published, and the sequence kicks off here, as a neurotoxin is released inside a private plane with horrifying results. If the scene works, and I believe it does, it’s largely because of the kind of tonal shift that I describe above. It opens with Powell and Chigorin discovering that there may be a lethal device on board the plane, and for several pages, the action unfolds like something out of a Tom Clancy novel, complete with detailed specs on the ventilation system. (The couple of paragraphs spent discussing the ram system and the mix manifold were the product of a lot of tedious hours paging through aircraft manuals online.) But once the poison is released, the tone shifts abruptly into nightmare, and the result is a page or two like nothing else in these novels.

In describing what Powell sees, I consciously turned back to the likes of King and Lovecraft, and there’s also a sentence or two of deliberate homage to “The Adventure of the Devil’s Foot,” a Sherlock Holmes short story that turns on a similar device. (“The Devil’s Foot” also provides the epigraph to Part III, and there are subtle allusions to it throughout the novel. Justice Roundhay, who sends Ilya to Belmarsh Prison, is named after one of Conan Doyle’s characters, and the two aliases that Karvonen uses—Dale Stern and Trevor Guinness—are nods to the names Sterndale and Tregennis.) The notion that Powell would see a monstrous version of one of the cherubim from Ezekiel’s vision of the merkabah is one of those ideas that seem obvious in retrospect, although it didn’t occur to me until fairly late in the process. It also involves a small cheat, since Powell is never directly privy to Wolfe’s conversations on the subject with Ilya, so I had to insert a short line in a previous chapter to explain why he’d have Ezekiel on his mind.

And although the result works well, at least to my eyes, I’m glad that it’s restricted to this chapter and nowhere else. Horror, as we all know well, is more effective the less it’s described, and as it stands, the description of Powell’s hallucination goes on just as long as necessary. It doesn’t feel like anything else in these books, which is part of the point: it’s a momentary disruption of the evenhanded tone I try to maintain even in scenes of great violence or intensity, and it casts a shadow over the more conventionally suspenseful scenes that follow. I’d love to write a real horror novel someday, mostly for the challenge of sustaining that kind of mood over a longer stretch of narrative: the number of novels that really pull it off would fill maybe a single shelf, and it’s no accident that King’s short stories are often so much scarier than his books. Still, I suspect that this scene works as well as it does because it’s embedded within a novel that otherwise seems so removed from the emotions that true horror evokes. And as with the poison that triggers these visions, a small dose is usually more than enough…

“So what are we saying here?”

Note: This post is the forty-fourth installment in my author’s commentary for City of Exiles, covering Chapter 43. You can read the earlier installments here. 

When you’re constructing an argument, whether in fiction, science, or philosophy, when it comes time to lay out your reasoning step by step, you’ll often end up presenting it in the reverse order from which it originally occurred. Sometimes a big idea will arise through accident, intuition, or the need to justify a preexisting position, and after taking it as a starting point, we search retrospectively for the evidence required to support it. Yet when we publish our findings, we pretend as if we arrived at the conclusion through a sequential chain of logic, proceeding from small to large, rather than the other way around. It’s clear, for instance, that scientists often proceed intuitively, at least when it comes to identifying a promising avenue of research, and that major discoveries can be the product of happenstance, luck, or trial and error. In most finished papers, that unruly process is reshaped into a tidy progression from hypothesis to conclusion, which seems only reasonable, given how deeply science depends on a shared language and set of standards. But it also skims over the mistakes, the dead ends, and the unquantifiable factors that affect any kind of intellectual activity.

We find much the same principle at work in fiction, most obviously in the mystery genre. While some authors, like Lawrence Block, can write most of a detective novel with only the vaguest idea of who might ultimately be responsible for the crime, most writers determine the identity of the guilty party early on, then go back to lay down a series of clues for the protagonist to follow. Unlike scientists, writers are careful not to make the progression seem too neat: it isn’t particularly satisfying to read a mystery in which every clue is handed to the hero on a silver platter, so a smart author builds in a few red herrings, wrong turns, and setbacks, an illusion of chance that has been as meticulously crafted as the solution’s apparent orderliness. In the end, though, the hero is moving through a series of encounters that the writer has put together in reverse, and if he often seems absurdly insightful, it’s only because he’s being steered on his way by an author who knows the ending. That’s equally true of puzzle mysteries, of the kind exemplified by Dan Brown, in which the protagonist is presented with a set of enigmas to solve. It doesn’t take much skill to come up with an anagram that the hero can crack at sight, but the reversed order of presentation makes it seem clever rather than a simple trick.

In the real world, this kind of backwards reasoning can be dangerous: reality is sufficiently dense and complicated that you can find evidence to support almost any theory, as long as you hide your work and present only a selective subset of all the available facts. One of the factors that makes conspiracy fiction so intriguing is the way in which it edges right up against the point of unforgivable distortion: unlike a pure mystery, the conspiracy novel takes real people and events as its material, but the result is little different from a whodunit that begins with the knowledge that the gardener did it, then scatters bloody gloves and shoemarks for the detective to find. In City of Exiles, for example, I make the case that the Dyatlov Pass incident—in which nine mountaineers died mysteriously in the snow in the Ural Mountains—was a test, conducted at the last minute, of a neurological weapon that was intended to bring down the plane of the Turkish prime minister Adnan Menderes. Menderes was, in fact, in a plane that crashed outside Gatwick Airport on February 17, 1959, or only two weeks after the Dyatlov Pass incident, and though he survived, the pilots were killed, and no convincing explanation for the accident has ever been found.

As presented in the novel, the links that join one incident to the other are structured as a chain of inferences leading inevitably to one conclusion. (It’s inevitable, at least, within the context of the story, although for reasons I’ve mentioned before, I was careful to pull back from the theory in the book’s penultimate chapter.) The writing process, however, was altogether different. I’d started with the Dyatlov Pass, and I knew for reasons of plot and symbolic resonance that I wanted to tie it into a plane crash—an image that occurred to me, in a totally arbitrary fashion, as a way to tie the story back to Ezekiel’s vision of the merkabah. With this in mind, I went back and searched for plane crashes that had taken place within the proper window of time, and the Gatwick accident had all the right elements. Everything else, including whatever motivation the Soviet security services might have had for killing Menderes, came after the fact. Laid in a straight line, it feels like it was conceived that way from the beginning, but it could have been very different. While doing my research, my attention was drawn to another plane crash that took place on February 3, 1959, just one day after the Dyatlov Pass incident. It was the crash that killed Buddy Holly. But if I’d gone in that direction, I don’t know how this novel would have looked…

“It took more courage to choose the passive way…”

Note: This post is the forty-third installment in my author’s commentary for City of Exiles, covering Chapter 42. You can read the earlier installments here. 

In theory, revision should be the easy part. You’ve written a complete draft of your novel, and you’ve set it aside for a few weeks or worked on another project or done whatever else you can to put yourself into a state of objectivity. After reading the whole thing over again, you go over it with a red pencil, keeping the good parts, cutting the bad stuff, and writing new material whenever necessary to hold it all together. When you’re finished, you’ve extracted a core of solid narrative from the untidy pages of the original manuscript, and although there’s still a lot of additional refinement to be done, you’ve survived the hardest part. I’ve said before, and I still mostly believe, that the difference between professional writers and amateurs often boils down to a willingness to cut: no book is so bad, as Pliny said, that there isn’t something good about it, and if you can bring out its positive elements while paring back the worst, you’ve already gone further than many other writers, who can’t bear to part with the labor of so many weeks or months.

Really, though, it isn’t that easy. Saying that revision is simply a matter of cutting the bad bits and keeping the good is a little like the old joke about how to make a sculpture of an elephant: you start with a block of marble, then take away anything that isn’t an elephant. It’s technically true, but in practice, you find yourself pushing up against all kinds of unanticipated problems. Maybe a subplot or a character or an entire narrative thread is fundamentally misconceived in ways that can’t simply be addressed by cutting what doesn’t work. Maybe the cuts you’ve made, while necessary, leave the novel feeling undercooked in other places, even after you’ve bulked up the rest. Maybe you just don’t have any idea how to make a certain essential scene on the page. Or maybe—and this is the most insidious issue of all—what’s good and bad in the draft can’t be easily separated. Our best impulses are often allied with the worst, and making cuts and adjustments in this case feels like pruning a tree than engaging in excruciating microsurgery.

I thought about this when I went back to reread Chapter 42 of City of Exiles, which feels now like it unites the book’s best and worst tendencies. It’s a simple scene, confined to Ilya’s prison cell, and the first half reads like the work of the kind of writer I’d like to be. It’s the first time we’ve been in Ilya’s head since he was beaten and locked in solitary confinement, but his mind, not surprisingly, seems to be on everything but his current predicament. In particular, he reflects on the binding of Isaac, one of the most mysterious scenes in the Old Testament, in which God tests Abraham’s faith by ordering him to kill his only son. The rabbis were clearly troubled by the story, and it inspired more exegetical speculation than any similar passage. Perhaps the most fascinating tradition—which I first encountered near the end of Gravity’s Rainbow—is that when Isaac was bound to the altar, he had a vision of the merkabah. And it stands here as an illustration of two contrasting approaches to wisdom: the active, restless kind, in which rabbis were described as storming the gates of heaven, and the far more difficult way of passivity and submission.

This is good stuff, and although it takes us away for a page or so from the main line of the narrative, I’m glad I kept it. (In any case, Ilya’s thoughts quickly take him back to the plot, and his reflections here provide the solution to the central mystery of the novel.) What happens next is less satisfying. Ilya, who knows that he needs to get word to Wolfe, acts in the only way he knows: when a guard comes with his evening meal, he overpowers him, taking him hostage, and uses him as a bargaining chip in exchange for a phone call. This isn’t bad, exactly, and I tried to make the action as plausible as I could, basing it on a similar incident involving the legendary British prisoner Charles Bronson. Still, when it read it again now, it strikes me as one of the instances in which this novel can be a little weak, falling back on prison movie conventions when I couldn’t find a way out of a particular scene. If I’d had more time, I might have been able to make it better, and it occurs to me now that I could have linked one moment—in which Ilya ties the guard’s hands with an improvised rope—more clearly to the binding of Isaac. Like Ilya, I often find myself taking the way of action instead of contemplation. And I wish I’d listened to my own advice…

“Most of the assembly was already done…”

Note: This post is the forty-second installment in my author’s commentary for City of Exiles, covering Chapter 41. You can read the earlier installments here. 

Hardware, as I’ve noted before, lies at the heart of a certain kind of thriller, and a lot of suspense novels seem to have written solely to showcase a particularly seductive bit of weaponry. (Two that come to mind, out of many possible examples, are Ken Follett’s The Hammer of Eden and Frederick Forsyth’s The Fourth Protocol, not to mention the Tom Clancy novel of your choice.) At times, the thriller comes startlingly close to science fiction in its fascination with technology, often in the form of gadgets and devices that don’t yet exist, at least not for the likes of us. In the Bond books and movies, hardware serves as another form of escapism, a sort of consumerist fantasy with Q as a combination of personal shopper and bespoke tailor. And even in superficially more realistic stories, technology feeds into the fantasy in a subtler way. An author’s familiarity with the details of guns or other tools of the trade grounds the more extravagant inventions of the plot, and we’re supposed to assume that if our writer knows what kind of holster would go best with a Walther PPK, he’s equally knowledgable about elements of spycraft and backroom politics that we have no way of verifying independently.

Of course, like all good narrative tricks, this one has its pitfalls, especially when the writer loses sight of the original intention. At its best, hardware can clarify and deepen a certain type of character: the heroes and villains of international suspense tend to be hypercompetent at what they do, even if they’re flawed in other ways, and we learn a little more about them as they go about handling their complicated equipment. All too often, though, technical details turn into an end in themselves, and we end up watching a name on the page take us through the fictional equivalent of a user’s manual. As with most descriptive or decorative elements, the amount a reader can tolerate is directly correlated to its apparent importance. When hardware isn’t essential to a particular plot point, the writer can, and should, get away with an evocative detail or two: an author like Thomas Harris, for instance, is a master at using bits of jargon or terminology to flesh out a passing moment. (“Lieutenant, it looks like he’s got two six-shot .38s. We heard three rounds fired and the dump pouches on the gun belts are still full, so he may just have nine left. Advise SWAT it’s +Ps jacketed hollowpoints. This guy favors the face.”)

When we’re dealing with an item of hardware that plays a more central role, we can indulge ourselves a bit more, and if we’ve handled it properly, the details enhance the story that follows: the object becomes a supporting actor in itself, and the action benefits in the same way in which a touch of backstory can enrich an important character. The ultimate example here is the rifle in The Day of the Jackal, which is more memorable than many of the human players involved—although it’s worth noting that we only care because it’s the weapon designed to assassinate Charles DeGaulle. On a more modest level, this also applies to the lethal device in City of Exiles. For most of the story, it’s a MacGuffin, designed only to push the characters from one violent appointment to the next, but as the climax nears, it becomes necessary to see exactly how it works. In Chapter 41, I devote a fair amount of time to describing how Karvonen puts it together, with particular emphasis on the cell phone detonator he constructs. All in all, it takes up about two pages at at point where the book has just over a hundred pages left to run, and I wouldn’t have sacrificed so much space to it if the effect hadn’t seemed worth it.

And there are a few distinct threads here. On the most basic level, I’d like to think that it creates a sense of anticipation: with every step in the process, we start to get a better idea of what this device is designed to do, even if the full details are withheld until the decisive moment. It gives us one last look at Karvonen as we’ve known him before, a careful craftsman, a few chapters before his plans start to spiral out of control. And it gives the reader just enough information to make the workings of this slightly implausible gadget more convincing. If I emphasize the detonator, rather than the heart of the weapon itself, it’s both because I didn’t feel entirely at home with the technical specs—which, thankfully, are hard to track down—and because I didn’t want or need to actually provide the reader with a handbook on building a particularly unpleasant device. In the end, Karvonen observes that the weapon isn’t exactly a thing of beauty, with three separate devices cobbled together with tape, but it works well enough for the task at hand. Which is more or less how I approached it in the writing process. It isn’t perfect, but it gets the job done. And we’re about to find out its true purpose…

“And this has something to do with Operation Pepel?”

Note: This post is the forty-first installment in my author’s commentary for City of Exiles, covering Chapter 40. You can read the earlier installments here. 

As I’ve written here elsewhere, research in fiction is less about factual accuracy than a way of dreaming. Fiction, like a dream, isn’t assembled out of nothing: it’s an assimilation and combination of elements that we’ve gathered in our everyday lives, in stories we hear from friends, in our reading and consumption of other works of art, and through the conscious investigation of whatever world we’ve decided to explore. This last component is perhaps the most crucial, and probably the least appreciated. Writers vary in the degree of novelistic attention they can bring to their surroundings at any one time, but most of us learn to dial it down: it’s both exhausting and a little unfair to life itself to constantly be mining for material. When we commence work on a project, though, our level of engagement rises correspondingly, to the point where we start seeing clues or messages everywhere we look. Research is really just a way of taking that urge for gleaning or bricolage and making it slightly more systematic, exposing ourselves to as many potential units of narrative as we can at a time when we’re especially tuned to such possibilities.

The primordial function of research—-of “furnishing and feathering a world,” in Anthony Lane’s memorable phrase—is especially striking when it comes to details that would never be noticed by the average reader. Few of us would care whether or not the fence at No. 7 Eccles Street could really be climbed by an ordinary man, but for James Joyce, it was important enough for him to write his aunt to confirm it. If we’re thinking only in terms of the effect on readers, this kind of meticulous accuracy can start to seem a little insane, but from the author’s point of view, it makes perfect sense. For most of the time we spend living with a novel, the only reader whose opinion matters is our own, and a lot of research consists of the author convincing himself that the story he’s describing could really have taken place. In order to lose ourselves in the fictional dream, the smallest elements have to seem persuasive to us, and even if a reader couldn’t be expected to know that we’ve fudged or invented a detail that we couldn’t verify elsewhere, we know it, and it subtly affects how deeply we can commit ourselves to the story we’re telling. A reader may never notice a minor dishonesty, but the writer will always remember it.

In my own fiction, I’ve tried to be as accurate as I can even in the smallest things. I keep a calendar of the major events in the story, and I do my best to square it with such matters as railway schedules, museum hours, and the times for sunrise and sunset. (As Robert Louis Stevenson wrote: “And how troublesome the moon is!”) I walk the locations of each scene whenever possible, counting off the steps and figuring out how long it would take a character to get from one point to another, and when I can’t go there in person, I spend a long time on Google Street View. It may seem like a lot of trouble, but it actually saves me work in the long run: being able to select useful details from a mass of existing material supplements the creative work that has to be done, and I’m always happier to take something intact from the real world than to have to invent it from scratch. And I take a kind of perverse pleasure in the knowledge that a reader wouldn’t consciously notice any of it. At best, these details serve as a kind of substratum for the visible events of the story, and tiny things add up to a narrative that is convincing in its broadest strokes. There’s no guarantee that such an approach will work, of course, but it’s hard to make anything work without it.

In City of Exiles, for instance, I briefly mention something called Operation Pepel, which is described as a special operation by Russian intelligence that occurred in Turkey in the sixties. Operation Pepel did, in fact, exist, even if we don’t know much about who was involved or what it was: I encountered it thanks to a passing reference, amounting to less than a sentence, in the monumental The Sword and the Shield by Christopher Andrew and Vasili Mitrokhin. (It caught my eye, incidentally, only because I’d already established that part of the story would center on an historical event involving Turkey, which is just another illustration of how parts of the research process can end up informing one another across far-flung spaces.) Later, I tie Operation Pepel—purely speculatively—to elements of the Soviet poison program, and the details I provide on such historical events as Project Bonfire are as accurate as I can make them. None of this will mean anything even to most specialists in the history of Russia, and I could easily have made up something that would have served just as well. But since I invent so much elsewhere, and so irresponsibly, it felt better to retain as many of the known facts I could. It may not matter to the reader, but it mattered a lot to me…

“We can’t trust our eyes or ears…”

Note: This post is the fortieth installment in my author’s commentary for City of Exiles, covering Chapter 39. You can read the earlier installments here. 

As I’ve noted here many times before, a conspiracy novel is really just an extreme manifestation of the rage for order that drives so much of fiction, as well as life itself. Given a seemingly random string of symbols, we’re naturally inclined to look for patterns, and the same holds true for events or artistic works that lend themselves to a range of interpretations. This is why paranoid readings of art and history so often go hand in hand. We tend to associate this inclination with the likes of Dan Brown, in whose books the reading of a painting becomes inextricable from a larger reinterpretation of historical events, but the impulse is much older and deeper. The urge to impose meaning on a text and to find a pattern in history, if not identical, are at least manifestations of a common need. And it’s no surprise that the methods used in both cases—analogy, juxtaposition, substitution, selective emphasis and deemphasis—are so similar. A conspiracy theorist poring over the records of the Kennedy assassination thinks in much the same way as a literary critic constructing a new reading of Pale Fire.

Yet there’s something qualitatively different about applying conspiratorial thinking to real history and doing the same to works of art. In the latter case, the reader runs the risk of distorting the author’s intentions and missing the work’s real value, but whatever harm it does is localized and subjective. A work of art should be open to various readings, and while some may be more valid than others, it’s easy to treat the process as a game. When we turn to actual events, though, the fallout from conspiratorial thinking is more troubling. Even in ambiguous situations, we know that there is one version of the truth, however hard it might be to uncover, and misrepresenting it does a disservice—or worse—to the facts. This is particularly true for events that occurred within living memory. When a theory began to circulate within days that the tragedy at Sandy Hook was a false-flag operation, we were rightly horrified, but few of us blink twice at stories that construct conspiracies around, say, Jack the Ripper, or even the Black Dahlia murder. And if we’re confronted by conspiracy theorists who pick targets that are too close to home, it’s tempting to respond, as Buzz Aldrin once did, with a punch to the face.

I like reading and writing conspiracy fiction as much as anyone else, but I’m uncomfortably aware of these issues. At the end of The Icon Thief, I was careful to blow up the paranoid story I’d constructed around Marcel Duchamp and the Rosicrucians, even though I’d like to believe that Duchamp himself would have been amused by it. City of Exiles posed similar problems. Like many conspiracy novels, it consists of two threads, one literary, focusing on the Book of Ezekiel, and one historical, focusing on the mystery of the Dyatlov Pass. When it came to the merkabah, I didn’t feel the need to hedge my bets: Ezekiel’s vision has been a locus for elaborate interpretation for centuries, and I felt that my reading—heavily indebted to David J. Halperin’s work in The Faces of the Chariot—was as valid as any other. The Dyatlov Pass was a different matter. This was a real event in which nine people died, and for those directly affected by it, the memory is a living one. I had what I thought was a plausible theory that covered much of the available evidence, but I wasn’t ready to commit to it altogether, especially because I suspected that many readers were encountering the story here for the first time.

In the end, I pulled back, although this doesn’t become clear until the novel’s closing pages. In Chapter 39, the two threads meet decisively for the first time, with Ilya and Wolfe moving toward a solution to both mysteries. Their appearance here together is no accident; throughout the novel, the study of the merkabah—which was said to call fire from heaven upon those who embarked on it without the proper preparation—has served as a metaphor for the investigation of secrets that might best be left in darkness. Here, at last, we also see that there’s another level of connection: just as a divine vision can lead to madness or death, the hikers at the Dyatlov Pass may have died in a similar way. Ilya only hints at the possibility here, and the full story will emerge gradually in the following chapters. The result is an extended piece of speculation and conjecture, and to my eyes, it’s at least as convincing as any other explanation that has been proposed. Ultimately, though, I undermine it, out of what I can only call respect for a tragedy that resists any definitive solution. I have a feeling that a lot of readers may have been left dissatisfied by this. But I really don’t think I had any other choice…

“He knew at once what was happening…”

Note: This post is the thirty-ninth installment in my author’s commentary for City of Exiles, covering Chapter 38. You can read the earlier installments here. 

In the workroom of the famed movie editor and sound designer Walter Murch, who remains one of my cultural heroes, there hangs a framed brass “B.” When asked what it means, Murch explains, in the paraphrase offered by Charles Koppelman in Behind the Seen:

Work hard to get the best grade you can—in this world, a B is all that is humanly attainable. One can be happy with that. Getting an A? That depends on good timing and the whims of the gods—it’s beyond your control. If you start to think that the gods are smiling, they will take your revenge. Keep your blade sharp. Make as good a film as you know how.

Ideally, we’d like to get the highest grade for every aspect of our work, but in reality, there will always be compromises. Part of the challenge of being a writer is focusing on what matters and refining those crucial parts until they reach their peak potential. Given constraints of time, energy, or space, this sometimes means that certain other elements will receive less attention. And while you always hope that any shortcuts you’ve taken will go unnoticed by the reader or be rendered invisible by distance, this is’t always true.

One of the hidden pitfalls of appreciating the importance of multiple revisions is the idea, or illusion, that you can fix any problems in the rewrite. Occasionally I’ll get stuck on some small point while writing—usually centering on how to get the characters from where they are now to where they need to be—and insert a makeshift solution for the first draft, trusting that if it doesn’t ring true on rereading, I’ll have plenty of time to come up with something better later. On the whole, it’s a good strategy; the alternative, as I’ve learned from hard experience, is an unfinished story, as you get hung up on solving any minor issues at hand before you go any further. I’ve argued many times that the answer to a seemingly unsolvable problem in Chapter 1 might become obvious by Chapter 20, but only if you’ve written the eighteen chapters in between, and I’ve generally done fairly well by trusting in the taste or ingenuity of the version of myself who will exist six months down the line. But sometimes I’ll go back and read something in print and wish I’d tried a little harder.

If there’s a sequence in City of Exiles that I’ve always thought was a little weak, it’s the section in Chapter 38 in which Ilya is attacked by another inmate at Belmarsh. The scene is there for sound narrative reasons: I’ve spent much of the novel building up the idea that Ilya has been thrown among enemies, and I couldn’t go to the end of the story without following through on those implications. The attack also occurs at the right place in the plot, at a moment when events are accelerating elsewhere and the other players have a strong reason for wanting to take Ilya down. That said, the execution of the moment itself isn’t particularly satisfying. The violence is borderline lurid, and it doesn’t offer much in the way of new ideas. Worst of all, Ilya’s attacker, a sinister inmate nicknamed Goat, is a nonentity. We don’t know much about him aside from a few ominous glimpses, and I didn’t take the time to work out who he was or where he came from. I try to come up with something a least a little distinctive even for minor characters, but I must have been distracted or careless here, so he doesn’t come off as a person so much as a plot point.

As a result, this is one corner of the story that I wouldn’t mind being given the chance to revise, although that ship has long since sailed. In my defense, I can plead all kinds of extenuating circumstances: I had about nine months to take this novel from outline to delivered draft, which remains by far the most compressed timeline I’ve ever had, and I was more concerned with delivering on the big moments. Even if the material here isn’t great, at least it’s over quickly, and it gets us to where we need to be in the next chapter. Of course, a reader isn’t particularly interested in how resourceful the novelist had to be to get a draft delivered on time; it’s only the finished work that counts. And I can’t help feeling that this could have been a big moment if I’d handled the buildup more capably. It’s possible that I’m exaggerating its shortcomings, and in practice, I suspect that most readers move past it without much thought either way. Still, I can’t help but see it as a lapse in what I otherwise think is a pretty good novel. Sometimes you just do what you can, take your B, and move on. And fortunately, it gets a lot better from here…

“He knew at last that he was home…”

Note: This post is the thirty-eighth installment in my author’s commentary for City of Exiles, covering Chapter 37. You can read the earlier installments here. 

I decided long ago that I didn’t want to make my background a central part of my writing. I’m Eurasian—of Finnish and Estonian descent on my mother’s side and Chinese on my father’s—but I’ve resisted, for better or worse, engaging directly with issues of multiracialism. Part of this is out of a dislike of being pigeonholed: there’s a lot of pressure in publishing to categorize authors, mostly as a way of differentiating them in the marketplace, and I don’t fault them for this. It’s hard for an author to stand out under any circumstances, and if your writing is grounded in a particular culture or milieu, it attracts readers who might otherwise have overlooked your work. But it also leads to an unfortunate situation in which an author of Asian descent, say, is expected to write primarily about Asian characters, exceptions like Kazuo Ishiguro notwithstanding. (None of this, by the way, should be taken as a slight against writers who have very good reasons for exploring their personal and cultural histories in fiction; it just isn’t what I ever saw myself as doing.)

As much as I welcome constraints of other kinds, this isn’t a limitation I wanted to accept, because I’ve always wanted to write about as wide a range of characters and situations as possible. (It’s worth noting, of course, that this attitude itself arises indirectly from my multiracial background, which long left me with a sense of not belonging to any group in particular, although that’s a subject for another post.) In practice, in much of my work, you’ll find hints of my background edging into the frame: Asian characters appear prominently in my short fiction, although usually in a way that doesn’t call attention to itself, and you could even argue that the dialogue between east and west embodied in my treatment of Russia is a way of approaching these issues in disguise. For the most part, though, I’ve written fiction that makes it very difficult to tell anything about my own ethnicity. I’m content to disappear from the story as much as I can, and although I’ve sometimes wondered if this has turned into a limitation of its own, it’s the mode in which I feel the most comfortable.

There’s one big exception, however, and that’s the treatment of Finland in City of Exiles. I’ve only visited Finland once, and Finnish-Americans are among the most highly assimilated groups in the United States, so the impact that this half of my background has had on my life has been relatively subtle. Yet I’ve always found Finland fascinating. It’s a liminal country, lying between Sweden on one end and Russia and the other, and its history has been fundamentally shaped by the presence of its enormous eastern neighbor. The population has never been large or diverse, which is a big part of the reason why it’s been so successful in building a strong welfare state, and although it’s since turned itself into a model nation, it sits only uneasily within the rest of Europe. Its language belongs to its own peculiar linguistic branch; its culture has been shaped by cold, darkness, isolation, war, and the legacy of the indigenous Sami people. And as remote as much of this experience has been from my own life, I still think about it whenever I glance at the sheathed puukko knife on my desk.

When I approached Finland as a location, then, it was both as an insider and an outsider, as well as a writer who saw it as a vein of interesting material that hadn’t been adequately explored in other novels. Karvonen first arrives in Helsinki in Chapter 37, and he regards it with a similar ambivalence: it’s a homecoming of sorts, but only for a man who has been estranged from his own country for a long time, to the point of collaborating with its historical enemy. We’re also introduced to the character of Laila, who plays a small but crucial role in the novel’s third act. Both are defined by their reactions to the memory of the Winter War, one of the bloodiest corners of World War II, and in writing two characters whose motivations are so rooted in history, it’s possible that I was belatedly discovering it for myself. I’m not sure it would have occurred to me to take Finland as a subject if it weren’t for my own family history, but when I look back, it remains one of the most satisfying memories of writing this novel. And if nothing else, I always knew that a Finn would make for a spectacular villain…

“He was visibly surprised to see the knife…”

Note: This post is the thirty-seventh installment in my author’s commentary for City of Exiles, covering Chapter 36. You can read the earlier installments here. Major spoilers follow for both The Icon Thief and City of Exiles.

I’ve spoken before of how tired I am of mechanical plot twists in suspense fiction, and particularly of how serial narratives, especially television shows, try to raise the stakes with the unexpected death of a major character. Of course, a thriller without a twist isn’t much of a thriller at all, and my objection has less to do with the quality or nature of any twist in itself than in its pathological overuse. The trouble with any trope that works is that writers tend to rely on it time and again, until, drained of all its original power, it settles into the status of a cliché. A real twist, as the term implies, should be a turning point, a moment in which the story takes on a permanently new direction, but in far too many novels, movies, and shows, it’s just business as usual, a continuation of a mode that leaves readers unsure of where they stand but saps the experience of much of its pleasure. Now that so many stories consist of twist after twist, we sometimes lose sight of the fact that one twist, judiciously employed, can be much more effective: good storytelling is about contrasts, and a twist gains much of its power from its juxtaposition with a narrative that, until then, seemed to be moving along more familiar lines.

In my case, it helps that the two most striking twists in The Icon Thief and City of Exiles—at least as measured by reader response—were both the unexpected product of necessity. Ethan’s death scene in the first novel was a very late revision in response to a note from my agent, who felt that the character’s original departure from the story, in the form of a suicide, wasn’t especially satisfying. Rewriting the story to give him a more dramatic sendoff required surprisingly few changes elsewhere in the novel, although it upset the balance of the narrative enough to result in a new epilogue that drastically altered the course of the series. With City of Exiles, as I’ve mentioned before, I didn’t know that Asthana was the mole until I’d already finished the first half of the book, which survives in its published form essentially unchanged. In both instances, I’d like to think that the fact that these were both unplanned additions increases their impact: if the reader is surprised, it’s probably because I was, too. (Speaking candidly, I have a feeling that the corresponding twist in Eternal Empire doesn’t work quite as well, if only because it was baked into the story from the earliest drafts.)

That said, a scene in which what seems like an ordinary conversation between two characters ends with one killing the other, in the first revelation his or her villainy, is a familiar one, so I had to work hard to make it feel fresh. The gold standard for such moments, as far as I’m concerned, is Jack’s valediction in L.A. Confidential, which has inspired countless imitations, from the sublime (Minority Report) to the workmanlike (24 and its successors, which have practically turned it into a tradition). If the version in City of Exiles works, it’s partially because the scene is written from Asthana’s point of view, putting the reader in her head as she feels increasingly uneasy around Garber, although the real reason for her wariness isn’t revealed until the last page. This kind of thing can feel like a bit of a cheat, and, frankly, it is. I played it as fair as I could, though, and while I essentially wrote the chapter as if Asthana were innocent and Garber a threat until that final turn, the logic holds up well enough on rereading. I don’t give the reader any false information; I just withhold a few crucial facts. And although this is an extreme example, it’s a familiar strategy in suspense fiction, which often relies on giving us only part of the picture.

While writing Chapter 36, I was also conscious that in its content and place in the story, it was uncomfortably close to the corresponding scene in The Icon Thief. I addressed this, first, by making the externals as different as I could. Along with narrating events from the killer’s point of view, I changed the setting—which is why much of the scene takes place in Garber’s car—and the murder weapon. (Asthana’s knife, a Spydero Harpy, is an obvious nod to its appearance in the novel Hannibal, and as the proud owner of one, I can testify that it’s a wickedly beautiful little tool.) I’m also pleased that it takes place in the shadow of the Battersea Power Station, one of the most striking buildings in London, best known from its appearance on the cover of Pink Floyd’s Animals. The choice of location was a pragmatic one: I wanted a secluded spot that was within a short drive from the Serious Organised Crime Agency’s headquarters in Vauxhall, and Battersea fit the bill perfectly. It also provides a resonant backdrop for Garber’s final speech about the centrality of energy to Russia’s political future, and how little the rest of the world can control it as long as it controls the flow of gas to Europe. I wrote those lines back in 2011, and in light of recent events, they seem even more true today…

“But the changes reveal more than they intend…”

Note: This post is the thirty-sixth installment in my author’s commentary for City of Exiles, covering Chapter 35. You can read the earlier installments here. 

Yesterday, I alluded to the cartographer Arthur H. Robinson’s story of how he developed his famous projection of the globe: he decided on the shapes he wanted for the continents first, then went back to figure out the underlying mathematics. Authors, of course, engage in this kind of inverted reasoning all the time. One of the peculiar things about a novel—and about most kinds of narrative art—is that while, with a few exceptions, it’s designed to be read in a linear fashion, the process of its conception is anything but straightforward. A writer may begin with a particular scene he wants to write, or, more commonly, a handful of such scenes, then assemble a cast of characters and an initial situation that will get him from one objective to the next. He can start with an outrageous plot twist and then, using the anthropic principle of fiction, set up the story so that the final surprise seems inevitable. Or he can take a handful of subjects or ideas he wants to explore and find a story that allows him to talk about them all. Once the process begins, it rarely proceeds straight from start to finish: it moves back and forth, circling back and advancing, and only in revision does the result begin to feel like all of a piece.

And I’ve learned that this tension between the nonlinear way a novel is conceived and the directional arrow of the narrative is a central element of creativity. (In many ways, it’s the reverse of visual art: a panting is built up one element at a time, only to be experienced all at once when finished, which leads to productive tensions and discoveries of its own.) In most stories, the range of options open to the characters grows increasingly narrow as the plot advances: the buildup of events and circumstance leaves the protagonist more and more constrained, whether it’s by a web of danger in a thriller or the slow reduction of personal freedom in a more realistic novel. That’s how suspense emerges, covertly or overtly; we read on to see how the characters will maneuver within the limits that the story has imposed. What ought to be less visible is the fact that the author has been operating under similar constraints from the very first page. He has some idea of where the story is going; he knows that certain incidents need to take place, rather than their hypothetical alternatives, to bring the characters to the turning points he’s envisioned; and this knowledge, combined with the need to conceal it, forces him to be more ingenious and resourceful than if he’d simply plowed ahead with no sense of what came next.

This is why I always set certain rules or goals for myself in advance of preparing a story, and it often helps if they’re a little bit arbitrary. When I started writing City of Exiles, for instance, I decided early on that the vision of Ezekiel would play a role in the plot, even if I didn’t know how. This is partially because I’d wanted to write something on the merkabah—the vision of the four fabulous creatures attending the chariot of God—for a long time, and I knew the material was rich and flexible enough to inform whatever novel I decided to write. More important, though, was my need for some kind of overriding constraint in the first place. Knowing a big element of the novel in advance served as a sort of machine for making choices: certain possibilities would suggest themselves over others, from the highest level to the lowest, and if I ever felt lost or got off track, I had an existing structure to guide me back to where I needed to be. And really, it could have been almost anything; as James Joyce said of the structure of Ulysses, it’s a bridge that can be blown up once the troops have gotten to the other side.  (Not every connective thread is created equal, of course. Using the same approach I’d used for my previous novels, I spent a long time trying to build Eternal Empire around the mystery of the Urim and Thummim, only to find that the logical connections I needed just weren’t there.)

Chapter 35 contains the longest extended discussion of Ezekiel’s vision in the novel so far, as Wolfe pays her second visit to Ilya in prison, and it provides an illustration in miniature of the problems I had to confront throughout the entire story. The material may be interesting in its own right, but if I can’t find ways of tying it back to events in the larger narrative, readers might well wonder what it’s doing here at all. (To be fair, some readers did have this reaction.) At various points in this chapter, you can see me, in the person of Wolfe, trying to bring the discussion back around to what is happening elsewhere in the story. According to the rabbis, Ezekiel’s vision can’t be discussed with a student under forty, and those who analyze the merkabah without the proper preparation run the risk of being burned alive by fire from heaven, which turns it into a metaphor for forbidden knowledge of any kind. And my own theory about the vision’s meaning, in which I’m highly indebted to David J. Halperin’s book The Faces of the Chariot, centers on the idea that elements of the story have been redacted or revised, which points to the acts of deception and erasure practiced by the Russian intelligence services. In the end, Wolfe leaves with a few precious hints, and if she’s able to put them to good use, that’s no accident. The entire story is designed to take her there…