Flickr user csessums // CC BY-SA 2.0 
Flickr user csessums // CC BY-SA 2.0 

What It's Like To Record the Audiobook Version of Infinite Jest

Flickr user csessums // CC BY-SA 2.0 
Flickr user csessums // CC BY-SA 2.0 

Internalized reading, like what you are doing right now, is a pretty comfortable pursuit. You have your own little narrator inside your head who probably sounds like you think you sound (or how you would like to sound), and you’ve likely developed a pretty good rapport with him or her over the years. You guys make a great team.

Attempt to read something aloud, however, and that quiet teamwork is torn to shreds. As a test, take the paragraph above and read it out loud. It’s not a difficult passage. There aren't any big words, foreign terms, or unruly surnames, though vocalizing it is still a chore. It's just three meager sentences, but your mouth is probably feeling a little dry. Now, think about reading a full page out loud. Daunting? Are you psychosomatically parched? Keep going. Imagine doing that again. And again. And again. Welcome to the life of an audiobook narrator—a voice inside your head for hire.

Sean Pratt has recorded over 800 audiobooks in his career. He also instructs aspiring narrators, offering a course on the ins and outs of the trade. “I’m teaching them about text analysis,” he tells me. “I’m teaching them about performance style, research, how to connect with the material.” His most important lesson may be what he tells prospective students before they start: “Narrate a book inside a closet for two weeks and then tell me if you’re still interested.”

Pratt’s niche in the industry is nonfiction, though he has done a little bit of everything, including a notable work of supposedly un-recordable fiction.


“I don't think my stuff's meant to be read out loud,” David Foster Wallace told an interviewer in 1997, a year after his novel Infinite Jest was published. For all practical purposes, he was right. The book is 1,079 pages of fractured narrative and teleporting endnotes, a complex meditation on addiction, tennis, depression, and media. How could one even begin to fathom reading all that out loud?

“It was a bit of a saga,” Pratt laughs. In 2009, he was approached by audiobook producer Hachette Audio to record Infinite Jest. At the time, he had never read it. “I don’t really read for pleasure anymore,” he says, noting that he records about 50 audiobooks a year. “I just don’t have time.”

The first thing Pratt noticed when Hachette sent him a formatted copy of Infinite Jest was, of course, the length. The second thing he noticed was the font, something you’re only really cognizant of if you read books out loud for a living. “They gave it to me in Tahoma, which is a sans-serif font. Sans-serifs are blocky, and they’re actually a little more difficult for me to read. When you couple that with Wallace’s idiomatic style of writing, it was going to be a real challenge.”

Here’s a small sampling of the kinds of things one can expect to do in order to narrate an audiobook version of Infinite Jest, based on Sean Pratt's experience:

1. Acquire dialect samples of French-Canadians to keep as reference material for extended sections of text pertaining to wheelchair-bound Quebecois separatists.
2. Assign and employ distinct speech patterns and affects to an entire cast of young boys and girls living at a tennis academy.
3. Open a dialogue with a NASA engineer in order to learn how to vocalize complex mathematical formulas. (Luckily, Pratt has a friend who is a rocket scientist.)

“It was the hardest book I’ve ever done," Pratt says, “but it’s one of my proudest.”

Without endnotes, the audiobook version of Infinite Jest ends up being around 56 hours. Because of his schedule, Pratt had to record it in sections, interspaced by the other books he was narrating at the time. He’d record an extended passage about, say, Boston’s grizzled Alcoholics Anonymous veterans (a.k.a. “Crocodiles”), then pivot to a romance novel or a physics text before returning to read one of Infinite Jest’s extended back-and-forths between a Quebecois separatist and a U.S. government agent in drag. Pratt can only narrate for about four hours a day before it starts to take a toll on his voice, so, in total, the Infinite Jest recording process took him about a year to complete.

The finished product is excellent and well worth a listen, even if it's just so you can hear how Pratt manages to do it. Wallace’s writing is famous for mimicking how our brains work, bridging tangential ideas like a mesh of neurons, dozens of these synapses firing in one sentence alone. “He starts out with idea number one, then he digresses to idea number two, then he goes to idea number three, then back to two, then back to three, then back to one, and back to three, and back to two,” Pratt says. “You have to vocally modulate that so the listener gets it.”

Keeping track of all that requires painstaking attention to detail. Before recording a section, Pratt marks the text like a musician scoring sheet music: "Breathe here … mark that … this is a digressive clause, so my voice needs to go down … this is a more important clause that needs to go up…”

The result is a clear, enjoyable narration of a supposedly un-narratable novel. Pratt occasionally gets fan emails about his Infinite Jest recording from admirers who have been reading the text along with him. “I didn’t realize it had such a hardcore following,” he says. “I feel like, in a way, it’s my own little piece of like being on Star Trek or something.”

At first, however, those hardcore fans were incensed—the original audiobook version lacked the novel's infamous endnotes.


Wallace's endnotes are purposely distracting. They are, in his words, “an intentional, programmatic part of Infinite Jest.” As he told the Boston Phoenix in 1998: “The way I think about things and experience things is not particularly linear, and it's not orderly, and it's not pyramidical, and there are a lot of loops.” For a reader, the physical act of flipping back and forth is mimetic of that process. For a listener, it’s not as effective.

Originally, Pratt didn't record any endnotes. Hachette instead decided to include a PDF with all that text for the listener to tote with them and read at the prompting of a British woman, who would interrupt the narration and chime in with the corresponding endnote number. "A lot of the hardcore Wallace fans were shocked and appalled," Pratt says. (“Why I Cannot Recommend the Infinite Jest Audio Book" was the title of a post published on The Howling Fantods, the largest Wallace resource site on the Internet, about the lack of endnotes.)

“There was no way in the technology to bounce back and forth,” Pratt says. “And there’s some stuff, like James O. Incandenza's filmography, that’s just pages and pages long. You run the risk of losing the narrative drive.”

Eventually Hachette relented, and asked Pratt to go back into his studio to record the endnotes. These add an additional 10 hours to the audiobook’s runtime, and you can download them as a separate file. It's up to you to find out how to flip back and forth between the recordings.

With endnotes added, the Infinite Jest audiobook nears 70 hours. Astonishingly, that's not even half as long as the longest book Pratt has ever narrated—a five-volume history of California that spans 150 hours. When asked if that text was tougher to record than Infinite Jest, he scoffs. "Not even close."


There's a tiny fellowship of people who know how difficult it is to sit in a room and read Infinite Jest out loud. Right now, their ranks stand at two. And, given the nature of copyright law, it looks to remain that way for a while.

In 1998, actor and audiobook narrator Steven Carpenter recorded his version of Infinite Jest, though you are unlikely to ever hear it. It was made for the National Library Service for the Blind and Physically Handicapped (NLS), and only participants of that program are allowed to listen. (Pratt has never heard it, but he and Carpenter are friendly and have commiserated about recording the novel.) 

Carpenter has narrated around 400 books for the service (which is run by the Library of Congress), and none of them will ever be released for commercial sale. "Anybody who is using this library, anyone who is blind or physically handicapped and wants to read those books is going to get to hear my interpretation of it," Carpenter tells me. "That's pretty cool. It also puts a burden on me, in a way. It's my responsibility to make it the best recording that I can."

Carpenter's version of Infinite Jest spans 10 cassettes (the Library of Congress went straight from cassettes to digital, skipping over CDs entirely), though he doesn't clearly remember the experience of recording it. "I know it was kind of a slog and there was certainly interest in the studio, with other people wondering how far I’d gotten and if I was finished yet," he says. Looking back, he estimates it took him about a month to record the text.

For his part, Carpenter recorded the endnotes as they appeared, meaning he skipped ahead and read them in their entirety whenever they popped up in the text. "There is one footnote in that book that is essentially a short story," he recalls, laughing. "I would say it was 10 or 15 pages, and all I could think while working on it was that the person who’s listening to this is never going to remember where we were when we left the bulk of the book to go down this rabbit hole."

Still, Infinite Jest isn't the most difficult audiobook Carpenter has ever recorded. That honor goes to William H. Gass's The Tunnel. "I didn’t know what I was getting into with that," he says. "The most difficult part of it is its stream-of-consciousness. It's very Joyce-ian. Gass doesn’t use quotation marks, so when there was dialogue I would have to read it through as carefully as I could, writing in the margins to suss out who is speaking. I'd put my notes in the margin for character one, character two, and so on, so I could map my way through it. And the run-on sentences," he groans, "or the not-even-complete sentences that go on for pages in some cases ..."

"Yeah," he laughs, "I’m kind of curious whether anyone has actually listened to it all the way through."

Hulton Archive/Getty Images
7 Lost and Rediscovered Literary Works by Famous Authors
F. Scott Fitzgerald
F. Scott Fitzgerald
Hulton Archive/Getty Images

A number of literary works by famous authors that were once thought lost have recently been rediscovered. Some were found in private collections, others within vast archives, and one was even uncovered in an attic. A few of these works have delighted readers and scholars alike, while others may have gone unpublished for a reason—yet all offer fresh insight into the development of the writers who wrote them.


In July 2015 Andrew Gulli, managing editor of The Strand magazine, was searching through the rare book archive at Princeton University when he uncovered a previously unpublished short story by Princeton alum F. Scott Fitzgerald. Gulli makes something of a habit of searching for lost and unpublished works by famous authors, and in the past has uncovered a story by John Steinbeck, which was also published for the first time in The Strand. Fitzgerald's 8000-word short story, entitled “Temperature” and written in 1939, features a hard-drinking writer with a heart problem. In a sad echo of real life, just a year after he wrote it Fitzgerald himself died of a heart attack.


Ted Geisel (Dr. Seuss) seated at a desk covered with his books
Library of Congress, Wikimedia // Public Domain

In 2013, the widow of Ted Geisel (better known as Dr. Seuss) rediscovered a pile of manuscripts and sketches that she had set aside shortly after her husband's death in 1991. The papers contained the words and illustrations for What Pet Shall I Get?, which was published by Random House in July 2015. It is thought the book was likely written between 1958 and 1962, since it features the same brother-and-sister characters found in Seuss’s 1960 bestseller One Fish Two Fish Red Fish Blue Fish.


Portrait of Sir Arthur Conan Doyle sitting at a table in his garden, Bignell Wood, New Forest, 1927
Fox Photos/Hulton Archive/Getty Images

A Sherlock Holmes short story supposedly written by Sir Arthur Conan Doyle was uncovered in the attic of historian Walter Elliot in 2015. The strange little story was written by Conan Doyle to be included in a collection of stories entitled The Book o' the Brig, which aimed to raise funds to rebuild a bridge across Ettrick Water, near Selkirk in Scotland, which had been destroyed during floods in 1902.

No sooner had the story been rediscovered, however, than some were expressing doubts about whether it had been written by Conan Doyle himself, especially since the flowery language doesn't seem in keeping with the renowned author's pared-down style. The full text of the story can be read (and puzzled over) here.


Photo of author Edith Wharton, wearing hat with a feather, coat with fur trim, and a fur muff
Beinecke Rare Books and Manuscript Library, Wikimedia // Public Domain

Alice Kelly, a researcher from Oxford University, was studying Edith Wharton’s papers in the Beinecke Library at Yale University in November 2015 when she discovered a previously unpublished short story. The unfinished nine-page story was stuck to the back of another manuscript, and is entitled "The Field of Honor." It centers on the First World War and is critical of the women who only superficially helped with the war effort, perhaps explaining why it was not published at such a sensitive time.


Crayon drawing of poet Percy Shelley circa 1820
Hulton Archive/Getty Images

When Percy Bysshe Shelley was in his first year of university at Oxford in 1810/11, he wrote and published a poem critical of the Napoleonic wars under the pseudonym “a gentlemen of the University of Oxford.” The 172-line poem was printed in a 20-page pamphlet entitled “Poetical Essay on the Existing State of Things” and was not attributed to Shelley until 50 years after his death. All copies were thought lost until 2006, when one was found amidst a mysterious private collection and offered for auction. Only scholars had access to the poem until 2015, when it was purchased by the Bodleian Library in Oxford to add to their world-famous collection of Shelley works and papers. The poem became the library’s 12 millionth book to be acquired and is now available online for all to read.


A black-and-white photo of a smiling Truman Capote
Evening Standard/Getty Images

A Swiss publisher poring over Truman Capote’s papers at the New York Public Library several years ago rediscovered a variety of short stories and poems the author had written before the age of 20. While four of the stories had been published in Capote’s school literary magazine, The Green Witch, the majority of the pile was brand-new to the reading public. In October 2015, Penguin books released the stories as The Early Stories of Truman Capote.


While looking through the archives of the city of Regensberg, Germany, researcher Erika Eichenseer uncovered 30 boxes containing more than 500 German fairy tales, which had lain unnoticed for 150 years. The stories had been collected by historian Franz Xaver von Schönwerth, who traveled around the Bavarian region of Oberpfalz recording folktales, myths, and legends in order to preserve them. He published the results of his research in three volumes between 1857 and 1859, but his matter-of-fact accounts of the stories were somewhat overshadowed by the more artful stories of his contemporaries the Brothers Grimm, and his book fell into obscurity. The Turnip Princess and Other Newly Discovered Fairy Tales contains 72 of the lost tales and was published by Penguin in February 2015.

A previous version of this story ran in 2015.

Oxford English Dictionary // Public Domain
The Time the Oxford English Dictionary Forgot a Word
Sir James Murray in his Scriptorium
Sir James Murray in his Scriptorium
Oxford English Dictionary // Public Domain

When the complete edition of what would become the Oxford English Dictionary debuted in 1928, it was lauded as a comprehensive collection of the English language, a glossary so vast—and so thorough—that no other reference book could ever exceed its detail or depth. In total, the project took seven decades to catalogue everything from A to Z, defining a total of 414,825 words. But in the eyes of its editor James Murray, the very first volume of the dictionary was something of an embarrassment: It was missing a word.

Looking back, it’s impressive that more words were not lost. Assembling the OED was a nightmare. Before the first volume—an installment consisting of words beginning with the letters A and B—was published in 1888, multiple editors had taken (and abandoned) the helm, and each regime change created new opportunities for mayhem. When James Murray took command in 1879, the Oxford English Dictionary could best be defined by the word disarray.

The irony of making this massive reference book was that it required millions upon millions of tiny, tiny pieces of paper. Every day, volunteers mailed in thousands of small strips of paper called “quotation slips.” On these slips, volunteers would copy a single sentence from a book, in hopes that this sentence could help illuminate a particular word’s meaning. (For example, the previous sentence might be a good example of the word illuminate. Volunteers would copy that sentence and mail it to Oxford’s editors, who would review it and compare the slip to others to highlight the word illuminate.)

The process helped Oxford’s editors study all of the shades of meaning expressed by a single word, but it was also tedious and messy. With thousands of slips pouring into the OED’s offices every day, things could often go wrong.

And they did.

Some papers were stuffed haphazardly into boxes or bags, where they gathered cobwebs and were forgotten. Words beginning with Pa went missing for 12 years, only to be recovered in County Cavan, Ireland, where somebody was using the papers as kindling. Slips for the letter G were nearly burned with somebody’s trash. In 1879, the entire letter H turned up in Italy. At one point, Murray opened a bag only to find a family of live mice chewing on the paperwork.

When Murray took over, he tried to right the ship. To better organize the project, he built a small building of corrugated iron called the “Scriptorium.” It resembled a sunken tool shed, but it was here—with the help of 1029 built-in pigeonholes—that Murray and his subeditors arranged, sorted, and filed more than a thousand incoming slips every day. Millions of quotations would pass through the Scriptorium, and hundreds of thousands of words would be neatly organized by Murray’s trusty team.

One word, however, slipped through the cracks.

Oxford English Dictionary entry slips
Oxford English Dictionary entry slips
Media Specialist, Flickr // CC BY-NC-SA 2.0

Bondmaid is not the kind of word people drop during conversation anymore, and that’s for the best: It means “a slave girl.” The word was most popular in the 16th century. Murray’s file for bondmaid, however, reached back even further: It included quotations as old as William Tyndale’s 1526 translation of the Bible.

But then bondmaid went missing. “Its slips had fallen down behind some books, and the editors had never noticed that it was gone,” writes Simon Winchester in The Meaning of Everything. When the first volume of the Oxford English Dictionary was published in 1888, bondmaid wasn’t there. (That volume of the OED does miss other words, but those exclusions were deliberate matters of editorial policy—bondmaid is the only word that the editors are known to have physically lost.)

When the slips were later rediscovered in the Scriptorium, Murray reportedly turned red with embarrassment. By 1901, some 14 years after the exclusion, he was still reeling over the mistake in a draft of a letter addressed to an anonymous contributor: “[N]ot one of the 30 people (at least) who saw the work at various stages between MS. and electrotyped pages noticed the omission. The phenomenon is absolutely inexplicable, and with our minute organization one would have said absolutely impossible; I hope also absolutely unparalleled.”

All was not lost for the lost word, however. In 1933, bondmaid made its Oxford dictionary debut. It had taken nearly five decades to make the correction.


More from mental floss studios