Wolfgang von Kempelen's Speaking Machine (1769)

From MSc Voice Technology
Jump to navigation Jump to search

Gourp Member: Yinqiu Wang/ YanhuaLiao / Ziyun Zhang / CantaoSu/ ChenyuLi

Introduction

In 1769, Hungarian inventor Wolfgang von Kempelen began developing a mechanical device capable of mimicking human speech, known as the "Speaking Machine"[1]. In 1791, Kempelen shared his insights and the inner workings of his "Speaking Machine" in a book titled "Mechanismus der Menschlichen Sprache" ("Mechanism of Human Speech and Language")[2]. His Speaking Machine consisted of a bellows to simulate the lungs, a reed to act as vocal folds, and a leather tube representing the vocal tract. By manipulating the shape of the tube, a variety of vowel sounds could be produced. Consonants were generated through the use of keys that controlled openings analogous to the lips, tongue, and palate.[3]


The Speaking Machine represented the first complete physical modeling of the human vocal system capable of synthesizing speech sounds.[4] Wolfgang von Kempelen’s pioneering work played a foundational role in the early exploration of artificial speech production. Beyond its significance in understanding vocal mechanisms, the Speaking Machine serves as an exemplar of human ingenuity in addressing complex challenges, even with the technological limitations of its time.[2] This introduction sets the stage for a comprehensive exploration of Wolfgang von Kempelen's remarkable contribution to the field of speech synthesis.

Historical Context

In the 18th century, there was a flourishing of mechanical engineering and automation, marking the era of automata, which also included advancements in speech automation technology. [5]Around 1780, Wolfgang von Kempelen produced the first ever fully functional mechanical speech synthesizer, known as the speaking machine. During this time, he became interested in the issue of speech among the deaf-mute community. More fundamentally, the prevailing importance and significance of speech motivated his work, reflecting the pervasive focus on this aspect of human communication during that period.[6]


In the century preceding von Kempelen, linguists, physicists, psychologists, and speech teachers engaged in extensive speculation and exploration of the mechanics of speech.[6]Meanwhile, the knowledge of sound, acoustics, and phonetics was relatively limited, so rather than basing his speaking machine on acoustic theories, Kempelen opted for an engineering approach of analysis-by-synthesis or trial and error. He aimed to achieve an audible result through a simple mechanism closely resembling our articulatory apparatus, yet playable like a musical instrument.[7]


However, the sounds produced by von Kempelen's synthesizer failed to meet the expectations of the 20th-century human ear, given the advancement in technology leading to the accustomed perfection or near-perfection in the reproduction of the human voice. Nevertheless, von Kempelen's contemporaries, unaccustomed to such technological advancements, were satisfied with the speaking machine, which adequately met their needs and expectations.[8]This era marked the initial explorations in speech technology, laying the groundwork for the subsequent development of sound synthesis and speech technologies.

Key Innovations

1. Mechanical simulation:

Kempelen's Speaking Machine garnered significant attention in its time because it was a mechanical device capable of mimicking human speech, which was considered a notable technological achievement in the 18th century. Although the technical principles of this machine significantly differ from modern speech synthesis technology, it served as an inspiration for subsequent research in speech synthesis and marked the early exploration of mechanical simulation of speech.[9]

2.Artificial Vocal Organ Simulation:

Kempelen's Speaking Machine included the simulation of human vocal cords, tongue, lips, and other sound-producing organs with mechanical components. While Kempelen was not the pioneer in this field, his Speaking Machine incorporated a more complex mechanical structure, which allowed for more precise simulations. Additionally, Kempelen's machine had the capability to control sound through mechanical devices, rather than merely passively imitating sounds. This was a significant technological achievement given the technological standards of the time.[10]

3.Multilingual support:

Multilingual support in Kempelen's Speaking Machine was a significant innovation with important historical implications. The machine possessed the capability to simulate different languages and dialects, which meant it could produce various sounds and pronunciations, not limited to a specific language. This was considered a technological achievement in the 18th century as it aimed to replicate the production of multiple speech sounds, thereby opening possibilities for research into different language communities and dialects.[11]

Impact

Von Kempelen's travail was not for naught and his impact far from nil.[4] Kempelen's speaking machine together with the book "Mechanismus der menschlichen Sprache" ("Mechanism of Human Speech and Language") he published in 1791 in which he described his observations on human speech production and his experiments with his speaking machine had a profound and multifaceted impact on the fields of linguistics and speech synthesis, influencing both the development of new ideas in linguistic research and advancements in audio engineering.

1. Contributions to linguistic research

While before Von Kempelen the larynx was considered central to speech production, his simple and successful demonstration drew the attention of 19th-century scientists to the vocal tract, the cavity between the glottis and the lips, as the main site of acoustic articulation.[12] In his book, he classifies the vowels according to her width of the lip channel giving a ranking of A >E>I>0>U and the width of the so called tongue channel that can be interpreted as horizontal tongue position. Although Kempelen isn't very explicit here, the observation clearly resembles the perceptual analysis of the second formant in whispered vowels described a century before by Reyher (1679; as cited in Kohler 2000; cf. Figure 12) and the vowel tunes of von Helmholtz (1862; cf. Figure 13). [13]This innovative vowel classification represented a novel approach to the study of speech sound physical properties.

2. Development of Speech Synthesis and Audio Engineering

Development of Speech Synthesis: Kempelen's research findings represented a crucial step in the development of speech synthesis. His insights into the resonating properties of the vocal tract and the impact of tube width on vowel production provided the foundational knowledge that subsequent researchers needed to simulate human speech mechanically.

Influence on Later Innovators: Later innovators built upon Kempelen's ideas, gained insights into resonating properties and developed theories of multiple resonance. In 1830, it was Willis, starting from the ideas of Kratzenstein and von Kempelen, who first gained resonable insight in the resonationg propgerties of neutral tubes that would be able to give the illusion of different vowels. And in 1838 Wheatstone who also rebuild Kempelen's machine added the theory of multiple resonance. [14] In about mid 1800's Charles Wheatstone constructed his famous version of von Kempelen's speaking machine. It was a bit more complicated and was capable to produce vowels and most of the consonant sounds. Some sound combinations and even full words were also possible to produce.[15]Inspirations in research and experiments with mechanical and semi-electrical analogs of vocal system in speech synthesis technologies were directly influenced by Kempelen's early work, making it a key milestone in the history of speech technology.

Future research

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. [16]

References

  1. Tarnóczy, T. H. “The Speaking Machine of Wolfgang von Kempelen.” The Journal of the Acoustical Society of America 21, no. 4_Supplement (July 1, 1949): 461–461. https://doi.org/10.1121/1.1917078.
  2. 2.0 2.1 Trouvain, Jürgen, and Fabian Brackhane. "Wolfgang von Kempelen’s speaking machine as an instrument for demonstration and research." In Proceedings of the 17th International. Congress of Phonetic Sciences. 17-21 August 2011, Hong Kong, pp. 164-167. 2011.
  3. Brackhane, Fabian, Jürgen Trouvain, and Richard W Sproat. “KEMPELEN Mechanismus der menschlichen Sprache,” n.d.
  4. 4.0 4.1 Schroeder, Manfred R. “A Brief History of Synthetic Speech.” Speech Communication 13, no. 1–2 (October 1993): 231–37. https://doi.org/10.1016/0167-6393(93)90074-U.
  5. Trouvain J, Brackhane F. The relevance of today Wolfgang von Kempelen's speaking machine[J]. Phonus. Research Reports of the Institute of Phonetics at the University of the Saarland, 2011, 16: 149-166.
  6. 6.0 6.1 Dudley H, Tarnoczy T H. The speaking machine of Wolfgang von Kempelen[J]. The Journal of the Acoustical Society of America, 1950, 22(2): 151-166.
  7. Pompino-Marschall B. Von Kempelen et al.: remarks on the history of articulatory-acoustic modelling[J]. ZAS Papers in Linguistics, 2005, 40: 145-159.
  8. van den Broecke M P R. Wolfgang von Kempelen's speaking machine as a performer[J]. Sound Structures. Studies for Antonie Cohen. Dordrecht: Foris, 1983: 9-19.
  9. Grassegger, Hans. “Von Kempelen and the Physiology of Speech Production,” n.d.[1]
  10. Trouvain, Jürgen, and Fabian Brackhane. “THE RELEVANCE TODAY OF WOLFGANG VON KEMPELEN’S ‘SPEAKING MACHINE,’” n.d. [2]
  11. Liénard, J.-S. (n.d.). FROM SPEAKING MACHINES TO SPEECH SYNTHESIS.[3]
  12. Schroeder, Manfred R. “A Brief History of Synthetic Speech.” Speech Communication 13, no. 1–2 (October 1993): 231–37. https://doi.org/10.1016/0167-6393(93)90074-U.
  13. Pompino-Marschall, Bernd. “Von Kempelen’s Contribution to the Theory of Acoustic Articulation,” n.d.
  14. Pompino-Marschall, Bernd. “Von Kempelen’s Contribution to the Theory of Acoustic Articulation,” n.d.
  15. http://research.spa.aalto.fi/publications/theses/lemmetty_mst/chap2.html
  16. Feng, S., Kudina, O., Halpern, B. M., & Scharenborg, O. (2021). Quantifying Bias in Automatic Speech Recognition (arXiv:2103.15122). arXiv. http://arxiv.org/abs/2103.15122