JOSEPH FABER SPENT SEVENTEEN YEARS working on a speech synthesizer in the mid-nineteenth century. When he finally perfected it, almost no one cared.
DECEMBER 22, 1845, A GERMAN IMMIGRANT NAMED JOSEPH FABER exhibited a mechanical device to the public at Philadelphia’s Musical Fund Hall. Outwardly it was similar to many curiosities of the era: A figure dressed like a Turk sat on a table, its face staring out at the crowd. The workings of this dummy, however, were far from the usual fare. Behind the figure a large bellows blew air into a series of complex mechanisms, seventeen years in the making, that imitated the human lungs and larynx along with important features of the mouth. As Faber—a balding, bespectacled man who spoke almost no English—played on foot pedals and a keyboard, the figure began to utter words, phrases, and even sentences in a weird, ghostly monotone. When the audience offered up phrases of its own and the machine repeated them, amazement turned to awe. This was no clever fraud but a genuine talking head.
Faber’s machine, disorienting in its time, is even harder to fathom today. Why would anyone spend so many years perfecting a device that, for all its ingenuity, had no apparent value except as a novelty? What motivated Faber, and what did he finally achieve for his efforts? And what did this strange mechanism mean to society? Did it lead to any other inventions, or was it simply a dead end?
The drive to create a mechanical double, whether it walks or talks or does the dishes, is very old. The notion that inanimate objects might somehow be made to speak dates back to early animistic religions and the myth of Pygmalion. Actual attempts to create such things are a different matter, of course, though by the time Faber’s marvel appeared at Musical Fund Hall, people had been trying their hands at talking machines for centuries.
The earliest speaking machines were perceived as the heretical works of magicians and thus as attempts to defy God. In the thirteenth century the philosopher Albertus Magnus is said to have created a head that could talk, only to see it destroyed by St. Thomas Aquinas, a former student of his, as an abomination. The English scientist-monk Roger Bacon seems to have produced one as well. That fakes were appearing in Europe in the late sixteenth and early seventeenth centuries is shown by Miguel de Cervantes’s description of a head that spoke to Don Quixote—with the help of a tube that led to the floor below. Like Magnus, this fictitious inventor also feared the judgment of religious authorities, though in his case he took it upon himself to destroy the heresy.
By the eighteenth century science had started to shed its connection to magic, and the problem of artificial speech was taken up by inventors of a more mechanical bent. In 1774 one Abbß Mical was said to have “solved the problem which had been believed insoluble” when he put a pair of talking heads on display in Paris. Around 1770 Friedrich von Knaus demonstrated a talking machine before the Austrian emperor and the duke of Tuscany. Little is known of their efforts. In fact, out of all the speaking-machine inventors of this period, only two left behind anything of note for historians to ponder. Luckily, the surviving documentation of these two inventors’ efforts is significant and provides a glimpse into how Faber developed his machine.
THE FIRST HARD EVIDENCE of workable talkingmachine technology comes to us from a man known simply as Kratzenstein. In 1779, when the Imperial Academy of Science in St. Petersburg, Russia, commissioned a contest to spur the study of speech production, Kratzenstein took the prize with a set of five tubes constructed in varying shapes. When a bellows blew air through the tubes, they produced the five basic vowel sounds.
At the same time, Baron Wolfgang Ritter von Kempelen, an accomplished inventor of automatons, was approaching the matter from a different angle. The main component of von Kempelen’s talking machines (he built at least two) was a hollow oval box, divided into two parts and fitted with jaw-like hinges. The first half of the box “caught” the air coming from the bellows, regulating the airflow as it passed into the second section, which functioned as a kind of mouth cavity. The jaws then rendered consonants as well as some crude differences in vowels.
Von Kempelen introduced other notable features, including a mechanism for mimicking nasal sounds. When a passageway to a pair of tin tubes was opened and the airflow to the jaws was shut off, the machine could produce a passable m sound. Opening only one of the tin tubes produced an n sound. And at least one of his machines had a keyboard, with about thirteen keys that operated four “jaws” and two “noses.” The result was an automaton, set on a table and covered with cloth, that could produce the letters a, o, u, p, l, and m and, as Goethe reported, could “say some childish words very nicely.” Von Kempelen himself did not consider it a success, however, and spoke of it only to a few close friends.
Exactly how Joseph Faber developed his talking machine is not known, nor is anything like a complete account of his life extant in English. He was born around 1800 in Freiburg, a university town near Germany’s Black Forest. He pursued a career in astronomy until failing eyesight diverted him into anatomy and mechanics. At some point he read von Kempelen’s book Über den Mechanismus der menschlichen Sprachen (On the Mechanisms of Human Speech). Some accounts say that he began working on his talking machine in 1829, while others suggest that he started as early as 1820. In any event, by 1835 he had gone far in synthesizing the findings of Kratzenstein and von Kempelen.
Expanding on von Kempelen’s arrangement, Faber built a rubber enclosure with two separate compartments. A bellows led into one side of this enclosure, where a small ivory reed, much like a clarinet reed, gave tone to the air. Inside the second half of the enclosure (a square funnel shape about two inches long) was a series of six sliding metal plates, which, like the Kratzenstein tubes, had variously shaped openings on their lower extremities. When these plates were raised or lowered, they created a finely nuanced current of air, allowing for pronunciations that had eluded von Kempelen. Even then it apparently took Faber seven years to teach his automaton to pronounce the vowel e .
THE REMAINDER OF FABER’S machine improved on other features established by von Kempelen. Passing out of the enclosure, the air reached the mouth cavity, which had evolved from a simple jaw hinge to a more anatomical affair with a tongue made of ivory and a palate, lower jaw, and cheeks made of rubber. In place of the tin tubes for nasal sounds, Faber fashioned a resonant chamber underneath the enclosure. Again drawing on von Kempelen, he connected the components to a series of levers ending in a keyboard, with sixteen keys for creating basic phonemes and a seventeenth for working a glottis. The final product was without a doubt the most complex speech mechanism ever built at the time. It could produce any vowel or consonant at will and could carry on a conversation in any European language.
And so, placing his device on a table and draping it in cloth, Faber set out to astonish the Continent. Unfortunately, the reaction of European audiences is not recorded, so the tale must pick up nine years later, when our mysterious German inventor set sail for the New World.
In the nineteenth century, as in the century before it, inventors often used live performances to publicize their accomplishments, and these shows often played on the gullibility of the crowd. Thus, when it came to inventions, the roots of ritual magic lived on as stage magic, sleight of hand, and misdirection. By contrast, scientists, who had recently escaped the stigma of alchemy, tried to distance themselves from such methods. American scientists in particular, having adopted many of the mores of the Puritans, preferred to circulate their ideas in a more dignified fashion.
It was the perennial conflict between high and low, so often irreconcilable, yet Faber’s machine was the rare case that spanned both worlds. Though Faber himself was more technician than thespian and generally made a point of revealing how his machine functioned, he seemed to imagine no place for himself beyond the stage. As a result, some were able to see his device as a work with serious potential, others as a boon to the box office, while a third group was able to see it as both. And as it continued to appear throughout the nineteenth century, it became emblematic of the changing times.
When Faber first exhibited his creation in New York City in early 1844, its future in the footlights did not look promising. A correspondent for the National Intelligencer interviewed the machine on February 29 and gave it high marks—“the only fault,” he wrote, “being a strong German accent, (which of course will wear off with travel).” Even so, as an exhibition curiosity it failed miserably. The draw was so poor that Faber moved to Philadelphia, only to find his reception there even worse. Soon, “in a fit of despair and intoxication” (a contemporary writer relates), he followed in the footsteps of his medieval forebears and “broke the image to pieces and burned up the fragments.”
Ironically, it was at this career low for Faber that other scientists began to wake up to the potential for artificial speech. Robert M. Patterson, director of the U.S. Mint in Philadelphia and for many years a prominent scientist, chanced upon Faber’s automaton a few days before its destruction and described it to the American Philosophical Society on May 17, 1844. The society’s members were sufficiently intrigued to recommend a charity drive to rebuild the machine. But Faber, still mourning, rejected the offer and repaired to his workshop, where he began reconstructing his device without any help from outside.
Among those who heard Patterson’s description that day was Joseph Henry, at the time a professor at Princeton, and soon to be the first director of the Smithsonian Institution. Henry had a complex relationship to both theater and technology. As a teenager he had run a theatrical company in Albany, New York, writing and producing his own plays and gaining the praise of the locals for his acting ability. Then at the age of sixteen he had turned his back on the stage to pursue a career in the sciences.
The decision proved to be a fruitful one. In 1831 Henry invented an electric generator, a few weeks before his British colleague Michael Faraday announced his own version. Soon afterward Henry became the first person to notice what we now call radio waves emanating from a live electrical wire. Later still, when Samuel F. B. Morse came to him for advice on building a better telegraph, Henry willingly obliged and (by his account) suggested crucial improvements, such as the use of relays and insulated wire.
IN DECEMBER 1845, AETER FABER had rebuilt his machine, Henry took time out from a scheduled visit to Philadelphia to see it firsthand. It was not the first talking machine he had ever laid eyes on. In April 1837 he had visited Charles Wheatstone, the British scientist who had designed a rudimentary machine based on von Kempelen’s version. “It … articulates with startling accuracy the words Papa Mamma Mother Father thumb plum and some other sounds,” wrote Henry in his diary. “It also laughs and cryes most admirably.”
Still, Henry had reason to be wary of Faber’s intentions. At the time, a device called the Anthropoglossos, or Mechanical Vocalist, was fooling audiences in a dimly lit room of London’s St. James Hall, thanks to the talent of a cockney ventriloquist off-stage. And of course, Henry had his reputation to think about. Eminent scientist that he was, he did not attend the show at Musical Fund Hall but went privately with Patterson to Faber’s workshop two days earlier.
At their personal demonstration the roles of man and machine were oddly reversed. “The German was studying the lesson for the exhibition,” wrote Henry afterward in a letter to Henry M. Alexander of Princeton; “he speaks but little English and Dr. Patterson was obliged to make him repeat the sentences which were to be said several times before they were properly articulated. With a little practice the figure really pronounced the words better than the operator[;] its organs were under more readily control than his own.”
Any fears Henry might have had of another Anthropoglossos were allayed during this visit. In fact, he used the moment to articulate his first vague formulations of the telephone. With such a device, he predicted, “words might be spoken at one end of the telegraph line, which had their origin at the other,” allowing preachers to sermonize from afar.
GIVEN HIS ROLE IN THE TELE graph, which Morse had unveiled a year and a half earlier, Henry could probably have developed an electromagnetic talking head for the nation’s clergymen. Perhaps his increasing interest in the pure sciences precluded such a venture. Perhaps he was still feeling the sting of his recent battle with Morse, who refused to credit Henry’s contributions to the telegraph and referred to his claims as “jackdaw dreams.” Whatever the reason, he never pursued the matter, and Faber’s talking machine embarked on a long career as a curiosity piece instead.
In 1846 Phineas T. Barnum, fresh from his success with the midget Tom Thumb, made a trip to Philadelphia looking for new material. Faber’s machine apparently filled the bill, because Barnum dubbed it the Euphonia and promptly packed it off to London along with sixteen Indians. By August Faber was demonstrating his invention, in conjunction with a musical organ he had built, to audiences at the Egyptian Hall, a popular entertainment venue of the time.
The dramaturgy in London remained much the same as it had been in Philadelphia. From its perch on a small table, the Turk, now swaddled in crimson and (despite its feminine features) sporting a full growth of beard, began by addressing the audience. “Please excuse my slow pronunciation,” said the voice. “Good morning, ladies and gentlemen. … It is a warm day. … It is a rainy day. … Buon giorno, signori .” After this salutation the crowd was invited to ask the figure to say whatever they wished. According to reports, it made the grade on such difficult words as thalassee, Lachlin Machlachlin, Mississippi, Massachusetts, Xerxes, Xenophon, and Xanthippe .
Accounts of the Euphonia’s success during this engagement differ. Many notables came to see it, including the Duke of Wellington, who offered his glowing endorsement. Barnum recorded receipts totaling three hundred dollars a week, with an expectation that the returns would increase “next year in the season .” Perhaps Barnum was simply putting on a good face, though, because most eyewitnesses recalled a small turnout.
One observant visitor, the future theater manager John Hollingshead, described his visit: “I paid my shilling and was shown into a larger room, half filled with boxes and timber, and badly lighted with lamps. In the centre was a box on the table, looking like a rough piano without legs and having two keyboards. This was surmounted by a half-length weird figure, rather bigger than a full-grown man, with an automaton head and a face looking more mysteriously vacant than such faces usually look. Its mouth was large, and opened like the jaws of Gorgibuster in the pantomime, disclosing artificial gums, teeth and all the organs of speech. … One keyboard, touched by the Professor, produced words which, slowly and deliberately in a hoarse sepulchral voice came from the mouth of the figure, as if from the depths of a tomb. It wanted little imagination to make the very few visitors believe that the figure contained an imprisoned human—or half human-being, bound to speak slowly when tormented from an unseen power outside. No one thought for a minute that they were being fooled. …”
Hollingshead’s account also offers a rare description of Faber himself. Gloomy and taciturn, he wore clothes that bore signs of the workshop, and his hair and beard “sadly wanted the attention of a barber.” At no point during his performance did he seem to acknowledge his audience. And when he contrived to make his device sing “God Save the Queen” for a finale,Hollingshead remarked that it “suggested, inevitably, God save the inventor.”
William Makepeace Thackeray, the Indian-born English novelist, was inclined to agree, although in satirizing the Euphonia in Punch , he unwittingly anticipated the phonograph: “A parson might set up the Compound Machine in his pulpit, and a clerk or curate work it from the reading-desk, whilst his Reverence was smoking his pipe in the vestry; or an under-secretary might set the bellows going with a speech of LORD JOHN’S whilst his Lordship was taking his usual glass of brandy-andwater at BELLAMY’S .”
BUT AT LEAST ONE VISITOR to the Egyptian Hall that summer regarded the talking machine as more than a passing fancy. At the time, Melville Bell, soon to become the father of Alexander Graham Bell, was developing a phonetic alphabet called Visible Speech that would be used to teach the deaf to speak. The Euphonia, with its mechanical production of speech, seemed to address the same concerns from an entirely different angle. So deep an impression did the Euphonia make on Melville Bell that it was still on his mind in 1863, when he took his sixteen-year-old son to meet Sir Charles Wheatstone.
Wheatstone, of course, had once shown Joseph Henry his own talking machine, and for the Bells he dragged it out again. The young Alec was clearly inspired by what he saw and heard that day. Not only did he borrow a copy of von Kempelen’s book from Wheatstone, but upon arriving home he convinced his brother, Melly, to join him in building a talking head of their own.
Alexander Graham Bell never banished drama and science to separate parts of his brain but instead allowed them to mingle and reinforce each other. This flexibility derived in part from his background. While his father tutored him in the field of phonetics, his grandfather inculcated a passion for the stage. By the time he met Wheatstone, he had already written a romantic comedy called Play of Douglas . (The original copy shows him working out symbols in Visible Speech in the margins.) Later on, even as he was experimenting with the telegraphic transmission of musical tones, his brother was disparaging “the extreme poverty of thrilling ideas which you manifest in most of your prose and dramatic works.”
The Bell automaton reflected the same catholic interests. On the mechanical side it had a bellows for lungs, a tin-tube throat, nasal cavities, a larynx, jaws, and a soft palate made of rubber stuffed with cotton. The tongue received special attention. Inspired by Faber’s device, it was cut into cross sections that could be raised or lowered at any point along its length.
To give this machine dramatic fullness, Bell planned to add a face and a wig. He probably would have gone through with it if his father had been more amenable to the idea. Certainly Alec reveled in the theatrical effect it created. When he worked the mouth and lips, the machine uttered the word mama with such clarity that his mother came running to find out what was wrong.
As an adult Alexander Graham Bell built on his early achievements, developing a “harmonic telegraph” and embarking on a career as a teacher of the deaf—projects that undoubtedly benefited from his experiments with talking machines. Faber was never as ambitious as that; for him a talking machine was an end in itself. Experience should have taught him that it was not much of an end, however. After his London engagement he toured the English countryside and again met with a lackluster response. Many people came to believe, erroneously, that he had destroyed both his machine and himself in some remote hamlet.
In fact, neither Faber nor the Euphonia perished in Britain. Barnum claimed to have bought the machine in the 1850s, though it was exhibited in France in 1862 without his name attached. Even when Faber did die, sometime in the 1860s, his creation did not vanish with him. It was passed down to his niece, whose husband fashioned himself “Professor Faber,” renamed the invention the Amazing Talking Machine, and took up the mission anew.
The Euphonia had undergone some alterations in the interim. Where once the head of a Turk had sat atop a robed half-figure, a woman’s head was now fixed to a scaffoldlike arrangement. The musical organ no longer provided accompaniment. The keys had been reduced from seventeen to fourteen, and the ivory tongue apparently replaced by one made of rubber. By this time there was also a special mechanism for creating a rolling r .
Yet for all the improvements, the cast of characters surrounding the Euphonia remained curiously unchanged. In December 1871 it appeared at Boston’s Horticultural Hall for a oneweek engagement, while Alexander Graham Bell was struggling with his telegraphic endeavors across town. Two years later it toured the United States in Barnum’s Traveling World’s Fair, with a new contingent of Indians, and stayed on for the next two seasons. In 1875, after these engagements had done little to fill the coffers, Professor Faber went to an aging Joseph Henry, its original endorser, and asked for financial support. Henry had once predicted a great future for the Euphonia as a serious work, but as a sideshow it held little appeal. He politely turned the entertainer away.
SO ONCE AGAIN THE AUTOMA ton was consigned to obscurity. In 1877, when Thomas Edison astounded the world with the phonograph, the Amazing Talking Machine received notice from Scientific American as a pale competitor—and understandably so. A “large affair as big as a parlor organ” was not so magnificent when a “little affair of a few pieces of metal” could achieve much the same thing. By 1879 it was in Paris, in a room adjoining the theater established by the great stage magician Robert-Houdin, where it was introduced without scientific authority and generally ignored.
Faber’s secret admirer, meanwhile, had advanced by leaps and bounds. In the same year that Professor Faber’s appeal to Henry went unheeded, Alexander Graham Bell arranged his own rendezvous with the venerable scientist to explain his frustrations in the realm of mechanical speech. Henry recognized Bell for the sincere man he was and implored him to learn as much about electricity as he could—in effect, to devise the electromagnetic talking head he had imagined three decades earlier. Bell took Henry’s words to heart, and in 1876 he developed a working telephone with Thomas Watson.
With Bell’s telephone, the conflict between the dramatists, who cared little for function, and the scientists, who cared little for publicity, was finally resolved. Bell and Watson went on to demonstrate their invention in live shows that combined concise technical explanations with brass bands, opera singers, and all the usual thrills of a theatrical performance. This approach had more than a passing impact. In proving that it was possible to be dramatic and still be honest, to entertain and yet produce a groundbreaking invention, the “telephone shows” set the tone for the next quarter-century, as America entered its justly named Golden Age of Invention.
As the 1870s turned into the 1880s, the Euphonia remained on exhibit in Paris, a relic of the past. Count Théodose Achille Louis Du Moncel saw it there and made a thorough investigation of its construction. His conclusions were laudatory. “When we consider the amount of study and experience which must have been applied to the combination of all these arrangements,” wrote Du Moncel, “it seems surprising that physicists have not given more attention to such an interesting machine.”
But at this point only one physicist thought the machine was still worthy of study. The Bell and Faber dynasties came face to face one last time in 1885, when Professor Faber approached Alexander Graham Bell for a loan of five hundred dollars. Bell, remembering his debt to the original Faber many years before, gladly extended his generosity. Then, four decades after its introduction at the Musical Fund Hall, the Amazing Talking Machine simply disappeared from the world stage.
THE TWENTIETH CENTURY HAS seen a number of successful attempts at speech synthesis using electronic methods. Bell Laboratories built machines called the Voder and the Vocoder, both of which give weight to the notion that the project stayed in the family. All that remains of the Euphonia, however, are the scattered accounts of a few witnesses. In the end the most amazing talking head of all was destroyed not by a priest, or even by its creator, but by the indifference of a world that had finally found its voice.