The Folklorist's Recorder

• Introduction
• Chapter 1 Etched in Phonograph Wax
• Chapter 2 The Acoustic Field Site
• Chapter 3 Listening to the Silent Page
• Chapter 4 The Horn in the Hollow
• Chapter 5 Capturing the Unscripted Voice
• Chapter 6 Aluminum Discs and Trunk-Sized Rigs
• Chapter 7 The Archive in the Automobile
• Chapter 8 Sounding the Color Line
• Chapter 9 The Illusion of Unmediated Performance
• Chapter 10 State Archives and National Sounds
• Chapter 11 The Revolution of Magnetic Tape
• Chapter 12 High-Fidelity Authenticity
• Chapter 13 The Pocket Collector: Cassette Culture
• Chapter 14 Rewinding the Storyteller
• Chapter 15 Noise Floors and Signal Isolation
• Chapter 16 The Visual Turn in Ethnography
• Chapter 17 DAT Tapes and Digital Fidelity
• Chapter 18 From Text Discipline to Media Science
• Chapter 19 The Spectrogram's Cold Eye
• Chapter 20 Ghost in the Machine: What Devices Leave Out
• Chapter 21 Who Owns the Field Recording?
• Chapter 22 Indigenous Sovereignty and Restricted Tapes
• Chapter 23 The Algorithmic Archive
• Chapter 24 AI and the Synthetic Folk Song
• Chapter 25 The Ethics of the Unrecorded Future

In the late summer of 1890, Jesse Walter Fewkes pointed a brown gutta-percha recording funnel toward a group of Passamaquoddy singers in Calais, Maine. When he engaged the clockwork motor of his Edison phonograph, a sapphire stylus began biting grooves into rotating cylinders of petroleum wax, cooling and curling off the mandrel like soft shavings of paraffin. Fewkes believed he was acting as a transparent medium—a modern transcriber whose apparatus could freeze the fugitive sound of an oral tradition before the forces of industrial modernity washed it away. Yet as the needle trembled, it did not merely transcribe the voice; it introduced a profound transformation into the nature of human memory. The songs that entered the bell of that mechanical horn emerged forever altered, separated from the bodies, the land, and the social contexts that gave them meaning, and reborn as durable, alienable artifacts.

For centuries prior, folklore was an art of the ear and the pen. Scholars like the Brothers Grimm or Francis James Child treated oral tradition as literature deferred—fleeting acoustic performances that attained intellectual legitimacy only when fixed as ink on paper. In this regime of the silent transcript, the collector wielded an absolute, often invisible editorial power: dialect was regularized, stumbles were smoothed over, and ribald verses were quietly expurgated to satisfy bourgeois reading tastes. The arrival of field recording technology broke that editorial monopoly, dragging the study of vernacular culture kicking and screaming into the laboratory. Suddenly, the stutter, the microtonal pitch drift, the intake of breath, and the barking dog in the farmyard were rendered indelibly audible. The apparatus demanded that collectors stop reading oral culture and begin measuring its acoustics.

The Folklorist's Recorder argues that field equipment was never an inert, neutral conduit for the preservation of culture; rather, the hardware itself actively rewrote the traditions it claimed to document. Every technological epoch—from the fragile wax cylinders of the Victorian era and the three-hundred-pound aluminum disc lathes hauled in the trunks of Depression-era folklorists, to the liberated mobility of quarter-inch magnetic tape, the cassette revolution, and today’s pristine digital file formats—imposed its own physical and aesthetic constraints on the informant. A singer performing into a phonograph horn was forced to crane their neck and shout, altering their natural dynamic range; a narrative that ran for three hours had to be butchered into three-minute bursts to accommodate the physical capacity of a seven-inch disc. Technology determined who could be recorded, what was preserved, and whose voices were deemed too erratic, too quiet, or too contaminated by background noise to qualify as "authentic."

By tracing this physical lineage of machines, this book reveals how the field transformed folklore from a branch of textual humanities into a contested media science. As folklorists traded notebooks for vacuum tubes and, later, oscilloscopes and spectral analysis software, their definition of authenticity moved away from the literary archetype toward the measurable fidelity of the signal. Yet this pursuit of scientific capture concealed deep, lingering ideological compromises. In the Jim Crow American South, across colonial outposts in sub-Saharan Africa, and across sovereign Indigenous reservations, the presence of the recording rig was fundamentally intertwined with dynamics of racial power, state surveillance, and institutional extraction. The collector possessed the battery pack and the export permit; the singer surrendered their voice to an archive they might never see, governed by intellectual property laws that treated traditional communal culture as uncopyrighted raw material waiting to be harvested.

Today, as magnetic tapes rot from hydrolysis, DAT cassettes demagnetize in basement archives, and centuries of ethnographic recordings are ingested as raw training data for generative artificial intelligence models, the battlefield has fundamentally shifted. The contemporary crisis of field recording is no longer about how to capture the voice, but who owns it once it enters the machine. Communities across the globe are fighting back against centuries of extractive archival practice, demanding data sovereignty, algorithmic repatriation, and the radical right to keep sacred or communal knowledge completely unrecorded. In an era where a machine-learning algorithm can ingest thousands of archival field tapes to synthesize an infinite supply of hyper-realistic, historically untethered "folk songs," we are forced to confront the haunting question the discipline has ignored for over a century: what has the machine been leaving out all along?

To look closely at the history of recording gear is to dismantle the pastoral romance of oral culture. This book offers readers an intimate, tactile history of folklore through the physical objects that mediated it—the zinc discs, the sticky tape splices, the wind-battered directional microphones, and the cold logic of the spectrogram. It invites you into the acoustic field site, where every microphone placement is an editorial decision, every background noise is a cultural artifact, and the hum of the mechanism is never entirely silent. By listening to the history of the apparatus itself, we begin to understand not just how traditional memory was preserved, but how the machine fundamentally reinvented what it means to remember.

In March 1890, Jesse Walter Fewkes stepped off a train in Boston carrying a heavy wooden shipping crate lined with excelsior wood shavings. Inside rested three dozen hollow cylinders of metallic soap and carnauba wax, their polished surfaces barely an eighth of an inch thick. To the baggage handlers on the platform, the cargo looked like overpriced tallow candles. To Fewkes, a marine zoologist who had pivoted with headlong enthusiasm into comparative anthropology, the dull yellow tubes were mechanical ledgers holding a living ledger of sound. They contained the voices of Peter Selmore and Noel Josephs, two Passamaquoddy men who had agreed, for three dollars a day, to lean over a table in a hotel parlor in Calais, Maine, and direct their voices into a cone of sheet zinc attached to an Edison Class M phonograph.

Fewkes was not the first human being to operate a sound-recording machine, but he was the first to realize that this particular corporate novelty could salvage an academic reputation. Thomas Edison had patented the phonograph thirteen years earlier, envisioning it primarily as an office dictation machine for high-powered executives, a singing doll for the nursery, or a parlor trick for curiosity-seekers willing to drop a nickel into a coin-operated slot. The early commercial phonograph was an unruly, heavy piece of office machinery designed for mahogany roll-top desks, powered by wet-cell chemical batteries that wept sulfuric acid and smelled of vinegar. It was not meant to cross mud ruts, survive salt fog, or sit on pine tables in remote border towns.

Fewkes borrowed the machine from Edison’s Boston office through an introduction arranged by Mary Hemenway, a wealthy philanthropist bankrolling the Hemenway Southwestern Archaeological Expedition. Hemenway wanted results, and Fewkes, whose previous academic work involved classifying jellyfish in the Mediterranean, was eager to prove that ethnology could be as rigorous, quantitative, and mechanically reproducible as marine biology. The notebook and the phonetic pencil, Fewkes wrote to his sponsors, were hopelessly flawed. When an ethnologist wrote down a tune, he heard it through the filter of his own musical training, squaring off strange quarter-tones into neat European intervals and sanding off the rough, guttering edges of unfamiliar vocal styles. The phonograph, by contrast, had no ear for harmony, no prejudice, and no memory; it simply cut whatever atmospheric disturbances hammered against its diaphragm.

The actual physical encounter between the Passamaquoddy singers and Fewkes’s leased apparatus was decidedly unglamorous. The Class M phonograph was driven by an electric motor hooked to a wet-cell battery that Fewkes had to keep charged using jars of toxic electrolyte solution, which sloshed menacingly whenever the cart jolted. The recording diaphragm itself was a sliver of mica glass glued to an arm holding a microscopic, chisel-shaped sapphire. To make the machine register an impression, Selmore and Josephs had to perform under conditions that violated nearly every social norm of natural singing. Sound waves do not possess much physical kinetic energy. To push the sapphire stylus through the resistance of the wax surface, the singer’s throat had to serve as an acoustic piston.

Fewkes shoved the narrow mouth of the funnel within an inch of Josephs’s lips. The singer had to yell. He had to keep his neck locked in place, because stepping back three inches to catch a breath dropped the signal beneath the noise floor of the machine, while leaning an inch too close caused the diaphragm to rattle against its brass mounting ring, yielding an unlistenable blast of distortion. Josephs sang a snake dance song, a war song, and counted from one to twenty, his face pressed against the cold zinc flange, his peripheral vision dominated by the spinning bronze mandrel revolving at a frantic one hundred and twenty revolutions per minute.

When the take was finished, Fewkes lifted the carriage, blew away the fine, silky curls of swarf—the debris carved out by the sapphire—and reversed the lead screw. He swapped the sharp cutting stylus for a rounded playback sapphire, attached a pair of rubber listening tubes resembling a doctor's stethoscope to the speaker horn, and handed the earpieces to his informants. When Josephs heard his own voice rattling back into his ears, tiny, metallic, and stripped of the natural room resonance of the parlor, he reportedly burst out laughing, then stared at the spinning mandrel with a mix of fascination and wariness. The voice had been detached from his lungs, cast in wax, and played back by a battery.

Fewkes was ecstatic. He mailed the cylinders to Boston, wrote up his findings for The American Naturalist, and declared that the machine had eliminated the subjective human observer from the documentation of vernacular culture. The phonograph, he insisted, preserved the absolute truth of the acoustic event. It did not lie, it did not forget, and it did not tire.

What Fewkes omitted from his published field reports was the degree to which the wax cylinder dictated what could and could not be preserved. The standard cylinder of 1890 was four inches long and two and an eighth inches in diameter. At normal recording speeds, it ran out of room in precisely two minutes and fifteen seconds. Any traditional ceremony, narrative song, or tribal council speech that unfolded over twenty minutes or four hours had to be brutally chopped down, accelerated, or artificially fragmented. Traditional storytellers, accustomed to cadence, silence, and interactive responses from an audience, were suddenly commanded by a white man with a pocket watch to begin on the count of three and wrap up before the brass carriage hit the stop-block at the far end of the mandrel.

The physical makeup of the wax was another silent arbiter of culture. The early blanks were not made of beeswax, despite the popular name, but of metallic soaps—lead, aluminum, or sodium salts mixed with stearic acid and ceresin, a petroleum derivative. If the blank was kept too cold, the wax became brittle, causing the sapphire to skip, chatter, and leave jagged, noisy tracks. If it was too warm, the medium softened into a gummy paste that clogged the cutting facet, muffling the high frequencies and smearing the sound into an unintelligible rumble. Fieldworkers in the American West had to warm the cylinders under blankets or over woodstoves before a session, praying that the wax would remain at an even consistency throughout the two-minute run.

A few months after his trial run in Maine, Fewkes took his Edison equipment to the desert Southwest, aiming the horn at the Hopi and Zuni pueblos. There, the desert heat introduced a new enemy: dust. The fine grit of the mesas settled on the oily surface of the blanks within seconds of unwrapping them. When the needle plowed into the groove, it dragged the sand along with it, creating a relentless frying-pan hiss that competed with the sacred chants. Worse still, the Hopi rituals were fundamentally communal and spatial. Dozens of dancers moved across plazas, their voices blending with the jingling of shells, the scrape of moccasins on rock, and the heavy thud of drums.

The Edison phonograph was completely blind to space. It was deaf to anything that occurred more than twenty-four inches from the mouth of the horn. If Fewkes placed the machine in the middle of a plaza, it registered only the thrum of the desert wind and a diffuse, useless murmur. To get a usable groove, he had to take individual priests and singers out of their ceremonial regalia, isolate them in a back room of the agency compound, and force them to shout sacred chants alone into a metal cone. The machine did not capture the ritual; it captured an artificial, high-pressure laboratory extraction of the ritual, performed under duress by an isolated informant who was stripped of his ceremonial community.

Yet within the burgeoning world of academic ethnology, Fewkes’s experiments set off a stampede toward the gadget. Anthropologist Franz Boas, then establishing his department at Columbia University, quickly recognized that the cylinder offered a powerful weapon against speculative armchair theorists. Boas had built his career arguing that human cultures must be studied through granular, objective empirical data rather than sweeping racial hierarchies. The phonograph looked like an ideal empirical engine. It turned intangible human breath into a physical, measurable commodity that could be shipped to a laboratory, measured with calipers, and preserved indefinitely in a museum cabinet.

In 1893, at the World’s Columbian Exposition in Chicago, Boas and his assistants gathered Indigenous performers from across the globe into a makeshift acoustic booth, running phonograph sessions alongside anthropometric skull-measuring stations. The cylinder was displayed as the sonic equivalent of the craniometer: an instrument that registered racial and cultural variance without literary embellishment. For the collectors, this was the dawn of salvage ethnology. The prevailing assumption among white scholars was that traditional cultures, particularly Indigenous North American tribes, were vanishing under the unstoppable advance of Western civilization. The phonograph was framed as an acoustic lifeboat, launched in the nick of time to salvage the vocal cargo of sinking cultures before they went under forever.

This panic over vanishing voices sparked a series of marathon recording expeditions that placed incredible physical demands on both the gear and the operators. Alice Cunningham Fletcher, an ethnologist working with the Bureau of American Ethnology, hauled an Edison machine into Nebraska to record the sacred ceremonies of the Omaha and Osage tribes, working in close collaboration with Francis La Flesche, an Omaha scholar who served as co-researcher and cultural intermediary. Fletcher had previously tried to transcribe Omaha songs by ear, employing a professional musician to arrange them for parlor piano, an enterprise that yielded absurdities like Omaha ritual chants fitted with polite European chord progressions. When she switched to the phonograph, she was startled to discover that the recordings exposed how much the piano had lied. The machine revealed rhythmic complexities that defied standard musical notation—microtonal shifts, polyrhythms, and guttural percussive accents that could not be scored on five lines and four spaces.

Yet Fletcher also ran headlong into the machine’s brutal physical fragility. Wax cylinders are soft. Every single time a rounded playback stylus traced the grooves to play back a song, it smoothed away a tiny fraction of the sound. After thirty plays, the highest frequencies—the crisp consonant sounds of the language, the rattle of a gourd, the high harmonic overtones of the human voice—were ground into silence. After fifty plays, the cylinder surrendered to a wall of gray acoustic noise.

Furthermore, the cylinders had to be shaved. In the 1890s, recording blanks cost upwards of fifty cents each—a substantial expense when an expedition's entire budget might be a few hundred dollars. To save money, collectors carried mechanical shaving attachments: small brass planes equipped with straight-edge blades that mounted directly onto the phonograph carriage. After transcribing a song onto paper or deciding that a take was flawed, the researcher turned a crank, and the blade sheared off a micro-thin layer of the wax, exposing a fresh, blank surface underneath.

Thousands of unique songs, prayers, and stories were systematically eradicated by the very folklorists who collected them, scraped off as swarf onto dirt floors to make room for the next informant. The physical archive was perpetually cannibalizing itself. If a collector ran out of fresh cylinders in the middle of a reservation hundreds of miles from a telegraph station, yesterday’s recording had to be sacrificed to capture today’s performance. The decision of what survived was often a logistical triage dictated by the bottom of the collector’s shipping trunk.

Even when preserved, the cylinders proved to be temperamental museum specimens. If packed in damp excelsior, mold spores took root in the animal fats and metallic stearates of the wax compound. The mycelium of the fungi spread through the microscopic grooves, eating the organic lubricants and leaving behind pitted, calcified tracks that snapped the sapphire needles of later conservators. If subjected to the dry heat of an attic or the bottom of an uninsulated trunk in an Indian agency basement, the paraffin separated, forming a white, crystalline bloom on the outer skin that obscured the recorded signal beneath a layer of chemical crust.

Beyond the chemical volatility of the medium lay an even deeper mechanical instability: speed. The early field phonographs were equipped with crude, spring-driven governors, or worse, unregulated electric motors tied to dying batteries. There was no standard, universal recording speed. A collector working in a hurry might record Noel Josephs at 100 revolutions per minute, Alice Fletcher might capture an Omaha pipe dance at 120 rpm, and a linguist documenting Arapaho verbs might crank the mandrel up to 160 rpm to improve the high-frequency reproduction of sibilant sounds.

To calibrate the speed for future listeners, the collector was supposed to blow a standard A-440 pitch pipe into the horn immediately before the singer began. Many collectors, rushing against an oncoming thunderstorm, a dying battery, or a restless singer, skipped this tedious calibration. As a consequence, generations of later musicologists playing back these cylinders had no reliable way of knowing the true pitch or tempo of the performance. If the cylinder was spun too fast, the singer sounded like an agitated child screeching on helium; if spun too slow, the performance dragged like a funeral dirge sung by a subterranean monster. The mechanical instability of the spring motor permanently warped the historical memory of how those traditions actually sounded.

The social transaction of the recording encounter was equally fraught. To the ethnologist, the phonograph was a neutral scientific listening post. To the communities being recorded, it was an unprecedented intrusion of industrial capital into sacred, private spaces. Sound, in many Indigenous cultures, was not simply aesthetic air vibrations; it was medicine, property, lineage, and power. Some songs belonged exclusively to specific clans, societies, or individuals, who had inherited them through generations or received them in dreams. To sing a winter hunting song in the middle of a hot summer was, in some traditions, an act of spiritual malpractice that could bring misfortune or physical sickness.

Yet the phonograph cared nothing for clan copyright, seasonal restrictions, or gendered taboos. It was a vacuum cleaner for sound. The ethnologist, desperate to fill his shipping crates before his research grant expired, pressured informants with cash, tobacco, and appeals to historical vanity. Some performers, realizing that the white visitor could not tell the difference, sang nonsense verses or bawdy jokes disguised as sacred liturgies. Others deliberately left out crucial verses, truncating the songs to neutralize their spiritual potency before feeding them into the zinc funnel. The machine caught the sound waves, but it was frequently duped by the singers, who used the physical limits of the cylinder as an acoustic shield to protect their real secrets.

The physical presence of the gear also dictated who could be a folklorist. An Edison field kit was not a casual purchase, nor was it light. A complete outfit—phonograph, metal horn, spare styluses, wet-cell batteries or heavy spring motors, a shaving attachment, and three crates of heavily padded, fragile blank cylinders—easily weighed more than one hundred pounds. It required pack mules, spring-mounted wagons, dry storage tents, and the disposable income of an elite institution. The physical difficulty of transporting wax cylinders into remote areas effectively excluded non-institutional collectors. Vernacular culture was documented almost exclusively by researchers who possessed the institutional backing, the technical training, and the imperial mobility to move heavy industrial hardware across domestic and foreign borders.

By the turn of the twentieth century, the phonograph had permanently altered the professional identity of the ethnographer. The discipline’s founding generation had been scholars of texts, men who spent their lives cross-referencing published collections of regional ballads or sifting through medieval manuscripts for traces of early myth. Their laboratory was the university library; their primary tool was the red pencil.

The new generation, raised on Edison wax, had to be machinists, electricians, and logistics coordinators. They had to know how to adjust the tension on a governing spring, how to hone the edge of a sapphire cutter using diamond dust, and how to mix a batch of battery acid without dissolving their own boots. The fieldworker was no longer just a listener writing notes in the dark at the back of a dance hall. They were the master of an acoustic spectacle, an engineer who parked a gleaming brass engine in the center of the village and ordered the community to rearrange itself around the needs of the revolving cylinder.

The sound that came back out of that horn was unlike anything humans had ever heard before. For thousands of years, oral tradition had been inseparable from the bodily presence of the human singer. A song died the instant it left the throat; it lived only in the communal memory of those who were present to receive it. Once etched into petroleum wax, the voice became a hard object. It could be dropped on the floor and smashed to pieces; it could be bought, sold, traded, and patented. It could outlive the throat that made it by decades, singing its two-minute song to rooms of strangers who did not know the singer's language, could not pronounce his name, and had never seen his face.

Jesse Walter Fewkes, sitting in his Boston laboratory late at night, would place the ear-tubes against his head and listen to Noel Josephs singing the Passamaquoddy snake dance, the mechanical rattle of the gear train humming along at a steady sixty cycles beneath the chant. The machine was noisy, the wax was fragile, and the voice sounded tiny and compressed, trapped inside a tiny acoustic cylinder. But it was there, permanently frozen in a thin ribbon of soap and paraffin, stripped of its life, but saved from the grave by the cold, indifferent bite of a sapphire chisel.