Choose Your Own Narrator

In one of my favorite Isaac Asimov essays, “The Ancient and the Ultimate,” Dr. Asimov writes considers a book and what would happen to it when combined with technological improvements over time. The essay was written in the early 1970s, but Asimov postulates that there would be electronic books, that the batteries would get better and better; that the book would remember where you last left off. Soon the book would no longer need a power source, and eventually, what you’d be holding in your hands is—well, a book!

I was thinking about this essay the other day while considering a similar mental extrapolation I was performing with audiobooks. I don’t recall Dr. Asimov mentioning audiobooks in his essay (I haven’t gone back to check), but as someone who consumes many of his books though the audiobook media, I wondered what the future holds for audiobooks.

Audiobook started out as “Books on Tape.” Already the medium has evolved quite aways from there. The first audiobook I ever listened to was Isaac Asimov’s abridged version of Foundation. It was two or three cassettes long, and I was annoyed by the fact that I had to change them.

Audiobooks moved from cassette to CD, and then, thankfully, to a digital format that allows them to be downloaded and stored on my iPhone, or even streamed over my laptop or desktop computer.

Like regular books, audiobooks used to be available only in bookstores. Now, from a place like Audible, I can have an audiobook within seconds. So what’s next for audiobooks? How might we imagine their evolution, in the way Dr. Asimov imagined the evolution of the book?

For me, the unique element of an audiobook is what the narrator brings to it. After all, the text of say of the book is same in paper, digital, or audiobook format (assuming all are the unabridged editions). A good narrator can bring a new dimension to that text, so it seemed to me that where audiobooks are most likely to evolve is with narration.

Consider that with an e-book, we can customize our experience somewhat. We can change the font type, or increase the font size; we can alter the brightness of the text to suit our preferences. We can turn a standard book into a large-font book, and vice-versa.

Now consider how the technology behind voice emulation is evolving. With tools like Alexa and Cortona, and Siri, we’ve gotten a pretty good grasp of creating unique voices from computers. So one possibility, going forward, is computer-generated narrators for audiobooks. Just like Siri, we might select the gender and accent of the voice that reads us a book.

That probably wouldn’t sit well with voice actors, however. So I see a second possibility. As the technology improves, it will be come possible for a computer, using machine learning, perhaps, to be able to perfectly emulate anyone’s voice, assuming they have enough of a sample to work off of. As it turns out, many of the best audiobook narrators have hundreds, if not thousands of hours of such digital samples already. If those samples could be used to produce a computer-generated emulation of a narrator, then it might be possible, someday, for audiobook listeners to be able to choose your own narrator—just another customization option.

Of course, there would need to be a mechanism for voice actors to be paid for this emulated use of their voice. Perhaps they would receive an additional royalty each time someone used their voice. Perhaps, it would be setup more like a ringtone, where you purchase and download voice actor emulations, and can apply them to any book you have. An app store for voice actor emulations would be available as part of the service offered by audiobook services like Audible.

Taking this a step further, it might become possible to do this for any voice. With permission from (and payment to) the estate of, say, Bing Crosby, an emulation of Bing’s voice could be created, and I could have Bing Crosby read me Neil Gaiman’s American Gods. Or better yet, I could have Isaac Asimov read me Shakespeare. Or Gilbert & Sullivan.

I’m optimistic that the technology won’t be the limited reagent in this particular evolutionary path. That will be left to the legal profession. But I suspect some kind of equitable arrangement can be reached.


  1. Interesting thought experiment.

    On the one hand, I think we’re still a long ways off. Good narrators won’t be so easily replaced by computers. At their best they are voice actors — they’re adding additional life and meaning to what they’re reading using different accents, inflections, emphasis, even the pauses they put between words. That’s all performance that enhances the story, but that doesn’t exist anywhere in the text. Yet, without it, you would get a pretty flat and lifeless result even if you could exactly model a particular narrator’s voice.

    (And here I’m also thinking about what’s going on with digitally created characters. It’s reached the point where we can do pretty amazing digital animation, but I still mainly see motion capture using real actors when they want to present the most life-like movements or facial emotions. I know the technology is there,, but the economics is that it’s easier and cheaper to capture that kind of human performance than to try to produce it from scratch digitally.)

    On the other hand, I suspect you’re right. We already routinely have limited conversations with Siri, Alexa, Onstar, voice mail systems,, etc. and can chose whether we want to hear things in American or British English, male or female (though I know some of those are “cheating” a bit by just combining snippets of pre-recorded phrases). Very basic text-to-speech systems are included in many operating systems now for accessibility reasons. It’s not that hard to imagine we’ll continue down those roads to a place where a computer can output any text with a very natural sounding voice *and* make be able to make better judgments than today about how best to present the materials based on the context.

    Maybe we’ll end up with basic readers that do a decent job and, like you said, will have the option to purchase “George Guidall” add-ons that gives a premium performance experience (just like you pay a little more for the 3D version of movies today at the theater).

    But before we solve things for audio-books, can’t we please, please, please do this for email? I’d give a ton for a good app for my phone that would read emails to me and give me decent hand-free controls to preview, skip, delete, etc. while I’m commuting in the car to work. You’d think these would have been out for years, but (at least in the Apple IOS world) I’ve yet to find anything that does even a decent job, including Siri. Listening to various apps try reading the full content of emails — a task where I’m not looking for a dramatic performance — sure suggests we have a ways to go yet 🙂

  2. I think you missed one additional advantage of modern audiobooks: bluetooth. Being able to connect your player to your car, or a myriad of other systems, means your story is always available. I love listening to audiobooks when I’m in thick traffic, as rock music makes me a bit to agitated and aggressive 😀 .

    The other amazing thing about audiobook narrators is how different they can be from one another. The first audiobooks I fell in love with were the classic Sherlock Holmes books. I recommended them to someone recently, who then asked me which narrator I preferred. Searching around, I struggled to find the voice I liked. I did not care for the most popular narrators, finding their reading to be too slow and inexpressive. It took me all day to find the voice I was familiar with, John Telfer (his reads can be found on YouTube and I strongly recommend them). It was surprising to see how different things can seem with different readings.


This site uses Akismet to reduce spam. Learn how your comment data is processed.