The Making of Early Literary Recordings
Jason Camlot




Early Phonographic Fiction

THE AUDIOBOOK AS we now think of it was not a material possibility in the early days of sound recording, but this did not inhibit speculation about such literary media artifacts.1 In 1878, soon after the introduction of the tinfoil phonograph, Edison forecast “a book of 40,000 words upon a single metal plate ten inches square” as “a strong probability,” adding that such phonographic books “would preserve more than the mental emanations of the brain of the author; and, as a bequest to future generations, they would be unequaled.”2 Others who witnessed the early demonstrations of the phonograph remarked on the significance of the new medium for the form and delivery of the novel on record. For example, Julian Hawthorne, who covered the literary beat for the journal America, speculated in 1888 on how the novel might be “produced on a sort of cooperative principle” when delivered with the technology of Charles Sumner Tainter’s graphophone (another manifestation of the wax-cylinder recording machine of the period). The novel, in this ideation, becomes a collaborative operation of audio production, with the author functioning as “a kind of stage manager and dramatist in one,” directing specialized actors and elocutionists in the enactment of her or his novelistic scenes and narration.3 Hawthorne’s vision of the audible novel as a full-fledged dramatization with personations and eloquently delivered narration was still a futuristic fantasy in 1888. Edison may have dreamed about having a novel in its entirety (he is said to have referred to Dickens’s Nicholas Nickleby as his example) on a compact audio record, but it was not until the 1930s, under the initiative of the Library of Congress Books for the Adult Blind project, that books the length of Victorian novels were transferred into the medium of sound.4 And even then, when Victor Hugo’s Les Misérables was produced in talking-book format on records that played at 331/3 rpm—much slower than the then commercial standard of 78 rpm—it still ran to an unwieldy 104 double-faced disks.5 The Edison cylinder and Victor flat disc record circa 1900 had limited affordances for the presentation of the novel as a genre.

This chapter examines some of the earliest adaptations of Victorian literary fiction into sound and focuses in particular on two extended examples that are useful for understanding how early spoken recordings were shaped by precipitant precedent media and forms of literary expression, and how audio technologies of the early twentieth century were imagined for use in teaching “new” kinds of literary experience. The story of early adaptations of Victorian fiction into sound introduces a variety of diverging plot-lines about remediation, all of which can be understood to display the twin logics of immediacy and hypermediacy—transparency and opacity—outlined by Jay Bolter and Richard Grusin, wherein “immediacy dictates that the medium should disappear and leave us in the presence of the thing represented” and yet simultaneously demands that the user take pleasure in the act of mediation by calling attention to the specificity of the new media form in itself and in relation to other media.6 The oscillation between immediacy and hypermediacy provides clues about how a new medium refashions older and other contemporary media, since the promise of “more immediate or authentic experience . . . inevitably leads us to become aware of the new medium as a medium.” Immediacy thus leads to hypermediacy.7

Sound-recording technology was marketed for its immediacy (among many other things) from the day it was introduced. The phonographic book was represented as a medium that would allow authors to communicate to future audiences with powerful immediacy and presence.8 The novelty of the medium lay in its capacity to store and reactivate the effects of authorial presence and literary immersion from one historical moment to another, rather than in the effect of vocalized authorial presence, itself. The voice of the author and storyteller was made available to the Victorian reading public “with all the living reality of the present moment,” repeatedly, in the form of “At Home” theatricals and public readings for decades prior to the existence of a talking machine.9 Any account of the new media claim for an invention like the phonograph—that it supersedes the print-based book in its delivery of vocal presence—must consider the Victorian book not as a silent repository of text awaiting an automated sounding technology but as the locus of what Ivan Kreilkamp refers to as Victorian “performative, mass reading,” understood as “a mode of literary consumption that is intersubjective, often occurring communally; vocal rather than silent; productive and active rather than passive and receptive; often occurring in public spaces rather than interior, domestic ones; and—perhaps most significantly—somatically responsive, involving a performance or display of physical reaction.”10

With this conception of reading in mind, Edison’s bequest might seem somewhat unidirectional and somatically delimited; it might sound inflexible, even a little tinny. It is precisely in such qualitative distinctions between mediated modes of literary expression that we can articulate a historicized conception of a medium’s relationship with a particular art form. To tell the story of early phonographic novels, both as they were imagined and as they existed, we need to consider the kinds of literary practice that informed them, as well as the literary works that furnished them with content to replay. In doing so, we come to understand the import and function of a medium and of the literary forms it can afford. John Guillory has made this point in a statement that serves to elucidate Bolter and Grusin’s argument about hypermediacy: “It is much easier to see what a medium does—the possibilities inherent in the material form of an art—when the same expressive or communicative contents are transposed from one medium to another. Remediation makes the medium as such visible.”11

Early spoken recordings were produced according to diverse models of generic adaptation, aesthetic and social value, display, dissemination, use, and experience. To recall Bauman’s terms, the way in which a recorded speech was entextualized, recontextualized, and performed had a significant impact on how it was understood and received by an audience. Approaching recorded performances for their “dynamics of recontextualization” reveals how specific early adaptations of Victorian novels into sound signified differently from each other.12 If you compare, for example, Len Spencer’s recordings made for Columbia Records in 1904 of “The Transformation Scene” and “The Flogging Scene” (based on The Strange Case of Dr. Jekyll and Mr. Hyde and Uncle Tom’s Cabin, respectively), “Svengali Mesmerizes Trilby,” as recorded by Herbert Beerbohm Tree for The Gramophone Company (later HMV) in 1906, and the numerous Dickens recordings made by Bransby Williams (between 1905 and 1948 for Edison, HMV, Columbia and other companies) and William Sterling Battis (for Victor in 1916), you encounter three different models of recontextualization and remediation.13

Spencer’s “The Transformation Scene” and “The Flogging Scene” recordings, dramatizations from popular novels, stand as the recordings of a seasoned phonograph performer playing roles on record that he had never played on the stage. It might be more relevant to understand Spencer’s recording of this transformation scene (which entails the performance of multiple character voices) in relation to his fame as a master mimic and monopolyloguist, than as a recording from a stage adaptation of R. L. Stevenson’s novella. This early adaptation of fiction into sound might be more usefully approached by considering phonographic reading formats than with reference to traditional literary genres. Spencer’s “literary” recording may well have been more recognizable as a descriptive sketch than as a fictional adaptation.

The meaning of “descriptive” as an audiotextual format, while wide-reaching, referred to a range of expected sonic elements, including the performance of multiple voices (sometimes by a single speaker) often of different speech dialects, enacting some kind of socially inflected scene (an auction, a journey, an encounter, a debate), with sound effects that both informed and were identifiable because of the framing scene, all delivered in a manner that conveyed spatial movement and depth through the strategic uses of sonic amplitude. Such records might have explicit verbal cues to situate the listener in the scene, or, might rely on sound effects to do that work for the recording. Spencer’s recording of “The Transformation Scene from Dr. Jekyll and Mr. Hyde” (1904/1908)14 opens with an announcement of the title (by Spencer, in a clear “American voice”) and then moves quickly into the refined Scottish brogue of Jekyll setting the scene with explicit contextualizing information, a sort of CliffsNotes leading up to the moment of transformation that will be portrayed in the recording. Here is a transcript of the recording in its entirety:

Jekyll: I have ransacked London in vain for the drug which has been the cause of all my misery. Half an hour after I have taken the last drop, [organ music] I shall sink into the terrible [unintelligible]. This, then, is the last time Henry Jekyll can see his own face, or think his own thoughts. I am losing my original self and becoming more and more incorporated with my worse nature. I go to sleep as Jekyll and wake up as Hyde. Will I die on the gallows, or will I have the courage to take this poison? What’s this I feel? The demon is coming. Hyde is– Ah, no, not yet, thank God. [Knocking.] What’s that? That way all arrive. Can it be my house steward? [Chiming bells.] Ah, those chimes. They remind me of that terrible night when, transformed as I was into that fiend incarnate Hyde, I murdered the father of the woman I loved. [Organ music.] Ah, I must pray—Pray God to keep away the demons. Ah, God, look into my heart and forgive my sins. You were right. I was wrong. Ah, ah the fiend is coming. Yes. Hyde is here!

Hyde: [Shrill throaty noises.] Stop that damned organ! The noise offends me ears! [Cackling laughter.] [Knocking.] They come for me! They’re going to take me to the gallows! [High tempo organ music.] But I don’t die on the gallows, oh no! [Cackling laughter.] I killed two people already. Here it goes for the third. Jekyll! Jekyll! I always told you I’d kill him. Ah, ah, aaooh [expiring cry].

Poole: [Knocking.] Break in the Door! [Sound of door being broken open.]

Utterson: Dr. Jekyll?! Why, it’s Hyde. [Organ music.] Dead.

Accompanied by mood-setting organ music as in melodrama, bell chimes that evoke the distant past as well as the acoustical space of the dwelling where the scene takes place, sound effects of door knocking and battering, and three distinct voices in different accents (those of Hyde, Jekyll, and Utterson the Lawyer ordering Mr. Poole, the butler to break down the door), this recording, while clearly the rendering of a scene from a stage adaptation of Stevenson’s novella, communicates all the elements that defined the “descriptive” audiotextual format in the context of early sound-recording culture. These generic elements of this particular kind of audiotext were developed to situate and render a concise series of events sonically. To make this point does not contradict the possible usefulness of the literary generic description of this recording as the dramatization of a work of fiction. But to apply that more traditional, print-oriented generic vocabulary does not fully capture the audiotextual format of the sonic artifact in question.

Tree’s “Svengali Mesmerizes Trilby” (1906) recording, on the other hand, can be understood as an adaptation from George du Maurier’s 1894 novel to Paul Potter’s popular dramatization, in which Tree performed the role of Svengali to great acclaim.15 The record was first advertised in the general HMV catalogue, and soon after listed in the Records of Unique and Historical Interest supplement to the general catalogue of 1910.16 Other artists whose recordings were included in this later, special catalogue included Sarah Bernhardt, Enrico Caruso, Leo Tolstoy, and Lewis Waller. The catalogue was designed to demonstrate “a new role” for the gramophone, given that by 1910 it had “been established sufficiently long” to undertake “the preservation of the art of a past generation.”17 In effect, the Beerbohm Tree recording was presented, almost from its first release, as a sound memento of historical significance. It was sold as one among several of the roles to which Tree applied his distinctive genius, along with Hamlet’s soliloquy on death, Falstaff’s speech on honor, and Mark Antony’s lament over the body of Julius Caesar, all of which were also recorded and released in that same year.

Listening to the recording itself, we hear that it was not the angelic singing voice of Trilby that was captured in the grooves of a ten-inch Black Label flat disc record but the “foreign”-accented voice of Svengali admiring the inside of Trilby’s mouth, fantasizing out loud about how he will dominate the world by using Trilby as a vocal instrument through which to communicate his musical genius, and then mesmerizing her for the first time. The bulk of the recording presents a sample of Svengali’s accent, which the novel describes as one that turns “a pretty language into an ugly one.”18 When Trilby’s voice is briefly heard (possibly rendered by the actress Dorothea Baird, who played Trilby in the Potter production, but we can’t know for sure), it is only her speaking voice, the voice of potential that Svengali is musically equipped to discover in her that we hear, and not the realized, ineffable voice of La Svengali singing under her master’s influence—not the vocal equivalent of Svengali’s cosmic piano. Considered in relation to the novel Trilby and its ubiquitous nostalgic impulses and themes, this recording represents the kind of memento that Taffy, The Laird, and Billee, the three musketeers of the brush who were infatuated with Trilby, would have loved to possess—a cylinder bearing the golden time when they were bohemian artists, when Trilby lived among them as some combination of mother, maid, and object of desire—to replay in the present. Or, more accurately, it represents the antithesis of such a souvenir, an anti-memento, insofar as it renders Svengali’s voice in lieu of the more transcendent sounds associated with that lost time. The 1906 Gramophone Company recording based on Trilby seems in a way designed to resist its status as event by capitalizing on compiling and communicating the fixed social stereotype of Svengali’s greedy, foreign-accented voice, rather than the protean voice of La Svengali as it is never fully captured in the text of the novel.

In many ways, the Beerbohm Tree recording is similar to the William Sterling Battis recordings based on minor characters from Dickens, in that it renders a paradigmatic event from du Maurier’s story in a manner that is abstracting and particularizing at the same time. Because the monologue provides the most typical verbal and active elements of Svengali by condensing a selection of his verbal ticks from speeches throughout the novel, and by making the speech itself an act of mesmerism, the recording works simultaneously as an abstraction of Svengali, a condensation of all of his most material eccentricities, and a speech act that, in the context of the larger drama of Trilby, was at the source of the sublime and protean voice of La Svengali, which can never be described or heard.19 In effect, this historical literary recording captures the inherent complexity (I am tempted to say impossibility) of the sound recording as a historical artifact. For all of our desire to capture the transcendent essence of past experience, to “preserve the art of a past generation,” to capture the sound of the magic that was, what remains instead in this adaptation of fiction into sound, is a grotesque caricature in dialect.


1. For a history of the modern full-length audio book, which begins in earnest with an account of the Library of Congress Books for the Blind project, see Matthew Rubery, The Untold Story of the Talking Book (Cambridge, MA: Harvard University Press, 2016).

2. Edison, “Phonograph and Its Future,” 534.

3. “It is very likely that when the human voice has taken its place in literature, companies of persons will exist whose calling will be to personate characters in stories and speak their dialogues. They must be persons of ability and culture, and marked histrionic ability. They must be, in fact, a higher class of actors. The novelist—who, in this case, would be a kind of stage manager and dramatist in one—would assign to each his part, explaining its nature and limitations, and the general purport of the situations and conversations. Then the actors must rehearse their parts together until, in the judgment of the manager, they were approximately perfect in them; after that the graphophones must come into play, and the story, in its final form, be talked into them by the dramatis personae. The narrative and descriptive parts could be given to a special elocutionist, trained for that especial function.” Julian Hawthorne, “The Human Voice in Literature,” America: A Journal of Today 1, no. 8 (May 26, 1888): 12.

4. “Edison estimates that Nicholas Nickleby can be transcribed upon six cylinders, six inches in diameter by twelve inches in length,” Philip G. Hubert Jr. reported in “The New Talking Machines,” Atlantic Monthly 63 (1889): 259.

5. John Cookson et al., Digital Talking Books: Planning for the Future (Washington, DC: Library of Congress, 1998), 3–4; Leroy Hughbanks, Talking Wax (n.p.: Hobson, 1945), 106. As Sterne notes, long-playing format records were first introduced in 1926 by Western Electric for the synchronization of sound with eleven-minute film reels. They were not used for the purpose of recording entire novels until over a decade later. Sterne, MP3, 14.

6. Jay David Bolter and Richard Grusin, Remediation: Understanding New Media (Cambridge, MA: MIT Press, 2000), 6, 14.

7. Ibid., 19.

8. “Julian Hawthorne in ‘America,’“24; Bleyer, 16.

9. See, e.g., Deborah Vlock, Dickens, Novel Reading, and the Victorian Popular Theatre (Cambridge: Cambridge University Press, 1998), and Kreilkamp, Voice and the Victorian Storyteller.

10. Kreilkamp, Voice and the Victorian Storyteller, 91.

11. John Guillory, “Genesis of the Media Concept,” Critical Inquiry 36 (Winter 2010): 321.

12. Bauman, World, 10.

13. “Transformation Scene from Dr. Jekyll and Mr. Hyde,” Len Spencer, performer, Columbia Phonograph Company 32604 (1905); “Svengali Mesmerizes Trilby,” Herbert Beerbohm Tree, performer, Gramophone Company, ten-inch disc, 1313 (1906); “The Awakening of Scrooge,” Bransby Williams, performer, Edison Company 13353 (1905).

14. Discography of American Historical Recordings, s.v. “Columbia matrix 1908. The transformation scene / Len Spencer,”

15. As the recording was described in the 1906 Columbia Phonograph Company Catalogue: “This tragic scene from the last act of the play depicts the final transformation of Dr. Jekyll into the demon Hyde, and his subsequent death by his own hand. The ringing of the chimes and pealing of the organ lend realism to the intensely thrilling climax.” Cited in Wurtzler, Electric Sounds, 350.

16. “Svengali Mesmerizes Trilby,” Herbert Beerbohm Tree, performer.

17. “The Art of a Past Generation,” His Master’s Voice: Records of Unique and Historical Interest (London: Gramophone Company, 1910), 2.

18. George du Maurier, Trilby (1895; Broadview Encore Edition, Peterborough, ON: Broadview Press, 2003), 28.

19. Alex Woloch, The One vs. The Many: Minor Characters and the Space of the Protagonist in the Novel (Princeton, NJ: Princeton University Press, 2004), 14.