The account which follows was first published in the New York Herald in the spring of 1868 and subsequently in a single volume that’s passed through several printings. It was the work of a professional correspondent of twenty years’ experience, and it shows. It’s well constructed. It’s readable.
It contains everything a responsible editor at a major American newspaper in 1868 would have permitted to appear in print, and nothing he wouldn’t.
This isn’t that account.
What this is, is the private record. The notes and observations Spilett kept alongside the published one during two and a half years on Lincoln Island and in the months after the rescue.
The published account serves the story. The private record serves the truth. They aren’t always the same thing.
This is one of my favorite premises of anything I’ve built on top of a classic. Verne handed me a narrator who’s a professional correspondent, and a correspondent always has a second set of notes. Once you notice Neb standing in the background of nearly every chapter with nothing going on inside him, you can’t stop seeing it, and I couldn’t leave him there. The way the original treats him is shabby, and the private record runs on that.
Why retell The Mysterious Island as a private journal?
Two reasons, and the first is simply how I write.
Very little of my fiction uses an outside narrator. It’s a standing preference instead of a decision I made for this book, and I find an interior viewpoint more interesting to write and more interesting to read.
The second reason is specific to a correspondent, and the book turns on it.
There’s always a difference between what a reporter files and what a reporter keeps in his notes. Always. There are things that don’t go in the public record, and the reasons vary, and none of them are recorded anywhere a reader can see.
Once you accept that, a Verne novel narrated from outside becomes something else entirely. It stops being the story and becomes the published version of the story. Then the obvious question arrives: what does the other version say.
What does a two-record premise do to a reader?
It changes what reading is, for the length of the book. A normal historical novel asks you to accept the narrator. This one asks you to hold two versions at once, and to notice the shape of the gap between them. Every scene arrives with a question attached: what did the Herald get, and why could this not go in.
That’s a demanding thing to ask, and it only works because the reader already knows the published account exists. Verne’s novel is famous. Most readers have some version of it in their heads, even if only the shape of it, and the prior knowledge holds up half the structure without my having to build it.
It also means the book can accuse without ever raising its voice. Spilett doesn’t denounce his younger self. He simply records what he wrote and what he knew, side by side, and lets a reader do the arithmetic. A man explaining his own omissions is more damning than any narrator condemning him would be.
What did Gideon Spilett leave out of the published account?
The entire story of Neb.
His name is Nebuchadnezzar, and he was called Neb. He was born in Cameroon and came to this continent by a passage that isn’t Spilett’s to describe. Like everybody else, he was on the island for the same two and a half years.
He was present in the published account on almost every page.
He’s absent from it entirely.
That’s the sentence the book is built around, and it’s exactly the mechanism by which a person disappears from a narrative while appearing constantly in it. Neb is there in every scene, doing things, being loyal, being useful. He’s never once in a sentence that matters. No interior life, no motive of his own, no moment where the narrative pauses to ask what he thought about any of it.
It isn’t done with malice. It happens through the ordinary operation of a narrative that knows, without being told, whose inner life is worth recording and whose isn’t.
I think that’s the worst kind of erasure, because nobody has to intend it. A writer who never asks whose head the story lets us into ends up repeating the prejudice of his century without noticing, and the person who pays is the character with no voice, along with every reader who recognizes himself in that character. Verne was a brilliant storyteller and he still did it. Plenty of writers working today do the same thing with far less excuse.
Why couldn’t the truth about Neb and Cyrus Harding be printed in 1868?
Because of what the relationship between Neb and Cyrus Harding was.
It was close. Closer than the published account allows, and equal in ways the published account can’t represent. It was a real bond between two men who had been through the same things.
In 1868, three years after the war, that was unpublishable.
Not the friendship in the abstract. The equality in it. A white engineer and a Black man on those terms, written plainly, wouldn’t have survived an editor at a major American newspaper, and Spilett knew it before he wrote a word. So he wrote a loyal servant instead. Devoted, capable, always present, entirely flat. The version the readership could accept.
That’s the thing the private record exists to correct, and correcting it required a book instead of a paragraph.
How do you write a character the original erased?
Carefully, and from outside.
I knew before I started that writing a Black Cameroonian’s interior life in 1865 was the hardest thing in this book. I did the historical research, the necessary part and the easy part to describe.
I knew before I started that writing a Black Cameroonian’s interior life in 1865 was the hardest thing in this book.Share on X
The harder decision was structural. I could have written the book from Neb’s point of view. That’s the obvious modern correction and there’s a version of this project that does it. I chose not to.
Writing from an exterior viewpoint, a white man observing during an extremely biased and prejudiced period, meant I didn’t need to go inside Neb’s head and declare what he was thinking. That gave me distance as a writer and a certain amount of leeway to be wrong, because the errors belong to Spilett’s perception instead of to a claim about Neb’s interior.
I think that’s more honest than the alternative. A white man in the twenty-first century writing a Black man’s inner life directly and presenting it as recovered truth is a larger claim than I’m able to make. A white correspondent in 1868 admitting he’s only the observable behavior and is inferring the rest is a claim I can stand behind.
What does labeling inference in a novel do?
Spilett marks his own guesses, and it’s the device I’m most pleased with.
He says outright that he had access to Neb’s observable behavior and almost nothing of his interior life. Neb kept that interior life with a discipline Spilett understood wasn’t incidental.
Where the record contains inference, it’s labeled as inference. Where a conversation has been reconstructed from something Spilett wasn’t present for, he says so. Where he’s speculating, he says he’s speculating.
That does two things at once.
It’s true to what a correspondent’s private notes would look like, because a working reporter of twenty years distinguishes reflexively between what he saw and what he assembled. And it keeps the book honest about its own limits. A reader is never told this is what Neb felt. A reader is told this is what one man observed and what he made of it, and the gap between those two is left open where it belongs. The discipline Neb kept is itself part of the answer. A man in his position in 1865 who let his interior show wouldn’t have survived the surviving. That he gave Spilett almost nothing is the record.
How is Cyrus Harding portrayed in The Mysterious Island?
In the original he’s a genius who’s always right. Engineer, leader, solver of every problem the island presents. The competence is real and I didn’t diminish it. What I added is the moral life underneath it, left unexamined in the published account.
He did something in the war. It wasn’t examined then and he’d prefer it not be examined now. The island doesn’t give him that option, because two and a half years with the same five people and nowhere to go isn’t an environment in which a man gets to keep a thing sealed.
The racism of the original stays in this version and gets examined.
Neb is still in the world that did what it did to him. Harding is still who he is. The gap between the published account and the record is, in part, the gap that race makes in deciding who gets to be a full person on the page.
I tried to close that gap without pretending it wasn’t there.
There’s a content warning on this book. Slavery, Civil War violence, and the racism of 1865, written as it was.
A version of this project sands the period down until a modern reader is comfortable. That version lies about the century. The prejudice in Verne’s original is the water everybody in 1865 was swimming in, and removing it would make the erasure of Neb inexplicable, since the erasure is a product of exactly that water.
Retellings that scrub the ugliness out of a period and call it sensitivity are a cop-out. A reader who never sees the prejudice can’t understand what it cost Neb to keep his interior to himself, and the writer who scrubbed it gets to feel good about a book that lies. I’d rather make a reader uncomfortable than let the century off the hook.
So it stays, and it gets examined instead of excused. Those are different operations and only one of them is worth a reader’s time.
Why does this retelling of The Mysterious Island add Magdalena Vasquez?
Because the story needs a woman.
Verne’s island has five men on it and no one else. That choice made sense to him and doesn’t make sense now. Magdalena is a New Mexican widow, she isn’t in the original at all, and she reads men like Harding faster than any of them read her.
What she does structurally is give the book somebody with no stake in the published version. The others were all there and all have reasons to prefer the story as told. She doesn’t, and a character with nothing invested in the official account is the most useful person you can put in a room.
Magdalena was the most fun I had writing this book. A woman who sees through Harding before anyone else does, dropped into a story that originally had no room for her at all, felt like opening a window in a room that had been shut for a hundred and fifty years.
How does The Mysterious Island fit with the other Jules Verne books?
Verne wrote three that connect, and I have now done all three.
20,000 Leagues Under the Sea is Nemo’s book. This one is the island. Eurydice is Ayrton’s confession.
The relationship between this book and Eurydice is closer than a reading order suggests. Eurydice overlays this one. The events of that book happen and run straight into this one, with real overlap between them, a consequence of Verne having written characters who cross between volumes and of my having taken that seriously instead of treating each as standalone.
The Time Machine sits alongside them in the same Victorian Reimagined group, though Wells stands on his own.
The architecture in all four is Victorian and it holds. What was wrong was the population.
Is Gideon Spilett a reliable narrator?
No, and he says so himself on the first pages.
He selected, shaped, and omitted in the published account with the full deliberateness of his profession, and he tells the reader he’s done it again here in different directions. He had access to Neb’s observable behavior and almost nothing of his interior life. Neb kept that interior life with a discipline Spilett understood wasn’t incidental.
So the record is labeled throughout. Where Spilett has reconstructed a conversation he wasn’t present for, he says so. Where he’s speculating about what somebody thought, he says he’s speculating.
A correspondent trying to correct what a correspondent did, and admitting he can’t fully manage it, is a more useful narrator than one who claims to have got it right.
How did five people end up on Lincoln Island?
A balloon out of Richmond in March 1865, and a question about which death was likelier.
Spilett’s framing of the decision is characteristic. The question was whether the attempt would kill them before the alternative did.
He’d reported from seventeen engagements of the war by then, stood in the telegraph office at Fredericksburg with grape-shot audible through the window, and twice been required to swim for his life. He records those facts and immediately says he isn’t recording them as boast.
On the precise sequence of events that made escape desirable, he declines to elaborate, because those events remain subject to the deliberations of parties whose interests he isn’t in a position to prejudice. That sentence is a correspondent telling you there’s a story he isn’t going to tell, in 1869, in writing.
Why does the published account still exist in The Mysterious Island?
Because the gap between the two records is the point.
The Herald version ran in the spring of 1868 and then in a single volume that’s been through several printings. It’s the account everybody read and it made the island famous, and it was the work of a professional correspondent doing his job well. Nothing in it is a lie. That’s what makes it useful as a comparison. A correspondent can select, shape and omit with complete professional integrity and still produce something that erases the most significant person present.
Spilett’s own summary of why he’s releasing the second record is the sentence the whole book runs on. The published account erases the most significant person on that island, not through malice, but through the ordinary operation of a narrative that knows without being told whose inner life is worth recording and whose isn’t. He’s describing his own profession. He’s also describing what Verne did in 1875, and he couldn’t have known that.
I’m prouder of this book than most of what I’ve done with Victorian material, because it refuses to let a beloved classic off the hook for who it left out. Verne’s island is still a great adventure. It also buried a man in plain sight on almost every page, and somebody should’ve dug him out a long time ago.
