Enquiry into language is to be found from the earliest times, and across many cultures and traditions. In some instances, such as in ancient India and the medieval Arab world, the study of speech (as distinct from language) was remarkably sophisticated. Nevertheless, it was in nineteenth century Europe that the modern field of phonetics became a clearly demarcated area of study. It grew up at the intersection of linguistic science with two other fields. One of them is mathematics and physical science, chiefly acoustics, and the other, medical science, especially physiology.

In Europe, nineteenth-century linguistic science itself was dominated by a comparative and historical paradigm, following the realization in the late eighteenth century that Sanskrit, Iranian, Greek, Latin, Germanic and Celtic were derived from a common parent language. Languages are modified over time by sound changes, which were understood as systematic modifications of pronunciation, and as a result languages which have evolved from a single parent language (which may itself be unattested) exhibit sound correspondences. The proper understanding of such regularities clearly requires a study of sounds, not of the written symbols which may be used to represent them.

There was a long tradition of articulatory classification of sounds (segments) according to the organs used to produce them, although some of the ideas were fanciful, and fundamental notions such as voicing and nasality were not clearly understood until the mid-nineteenth century. The century brought advances in scientific medicine and physiology, and in 1856 Ernst Wilhelm von Brücke published a treatment of speech physiology and sound classification intended for linguists and teachers of the deaf which set articulatory phonetics on a secure foundation.

The eighteenth century saw attempts at speech synthesis, culminating in von Kempelen’s bellows-powered speaking machine of 1791. In 1830, Robert Willis published his experiments showing that a range of vowel-like sounds could be produced by using a vibrating reed to excite the resonances of a tube of variable length. In 1863 Helmholtz published a seminal work (translated into English as Sensations of Tone) which is a comprehensive treatment of sound, covering analysis, synthesis, and hearing, and taking into account the sounds of speech. By the 1880s it was understood that the voice provided an excitation rich in harmonics and that vowel sounds differ one from another according to the reinforced regions (formants) in their spectra.

Instruments began to appear which could make enduring registrations of fleeting sounds. The phonautograph was the first primitive oscillograph which rendered the speech waveform visible on paper, and in 1877 Thomas Edison announced the phonograph which permitted the recording and reproduction of speech. Only a year later, vowel waveforms obtained from enlarged phonograph grooves were subjected to quantitative harmonic analysis by Jenkin and Ewing, thus anticipating the way in which speech signals are analysed today.

The insight that a finite phonetic alphabet of reasonable size might be established which could represent all the sounds to be found in any language was clearly formulated at least as early as 1796 by Thomas Young. The feasibility and development of such an alphabet was the original subject of the Prix Volney, a prestigious annual essay prize which was awarded from 1822 onwards. Colonial expansion, together with Christian missionary work, greatly extended the range and diversity of languages to be dealt with.

By the mid-nineteenth century, the term “phonetics” had come into use in English. Descriptive practical phonetics and notation underwent its most extensive development in Britain. Alexander Melville Bell developed “Visible Speech”, a type of organic alphabet in which the invented letter shapes directly represent the glottal state and articulatory configuration. But eventually alphabets based on the roman alphabet gained the ascendancy. Alexander John Ellis and Henry Sweet made fundamental contributions to phonetic terminology and classification.

The International Phonetic Association (IPA) had its origin as The Phonetic Teachers’ Association formed in 1886 by a small group of teachers of English in Paris headed by Paul Passy. They were inspired by a new trend in modern language teaching, The Reform Movement, which launched in 1882 with a manifesto by Wilhelm Viëtor. At first the association used already existing phonetic transcription systems. In 1897 the group was renamed the International Phonetic Association and began to be concerned with wider theoretical issues. Eventually the development of an International Phonetic Alphabet (also IPA) was adopted as one of the association’s goals. The association’s journal, originally published in phonetic script, was renamed in 1971 as Journal of the International Phonetic Association when publication shifted to conventional spelling. It is the longest established journal in the field. For much of the twentieth century, the IPA was sustained by an administrative and editorial hub at University College London (UCL), and Daniel Jones, the first head of the Phonetics Department there, played a large part in the development of the alphabet. The development of the alphabet was guided by the phoneme theory (see Ashby and Maidment (2005), Chapter 9), to which Jones also made contributions.

The techniques and instruments of experimental phonetics began to be settled from about 1890 onwards, particularly under the leadership of the abbé Rousselot in Paris. His dissertation, widely hailed as ground-breaking, appeared in 1891, the first volume of his Principes de phonétique expérimentale was published in 1897, and he set up a dedicated laboratory of experimental phonetics at the Collège de France in 1898. ‘Experimental phonetics’ became a craze, and suddenly universities everywhere wanted the prestige of a phonetics laboratory. By 1916 there were already about 25 phonetics laboratories around the world, mostly in Europe.

Of course, there had been previous ‘experimental’ and instrumental work on aspects of speech, and some of this had been conducted in ‘laboratories’ but they had been physiology or psychology laboratories. The new kind of phonetics laboratory modelled after Rousselot is a dedicated facility, independent, pursuing its own agenda, and maybe even employing technical staff. Its task was to provide a whole range of instruments and methods—especially the two cornerstone techniques of Rousselot’s approach which were palatography and kymography.

The first announcement of the technique we now know a palatography was made by an English dentist, James Oakley Coles, in 1872. The method of investigation devised by Coles was direct palatography using a mixture of flour and gum applied to all the upper surfaces of the mouth—teeth, hard and soft palate. After articulating what he called a single ‘letter’, he observed the wipe-off pattern. The indirect method of palatography, using an artificial palate which could be removed for examination, was described very soon afterwards by the American Norman William Kingsley, another dentist. The indirect method became the norm for the first half of the 20th century.

Awareness of the palatographic method entered the linguistic-phonetic mainstream via the second edition of an important handbook of Sievers in 1881, and palatograms were used to illustrate many late nineteenth-century and early twentieth century textbooks.

Throughout the period from 1890 up until the 1950s, the large recording drum of the kymograph was the most conspicuous piece of machinery visible in any phonetics laboratory. Various pressure transducers (typically nose, mouth and larynx) are connected by air-hoses to capsules sealed by flexible membranes. Movement is transmitted to lightweight levers which scratch traces on soot-blackened paper moving on the drum. A mask over the mouth collects air flow, as well as acting as the input to an acoustic speaking tube. The resulting wave is an unsystematic blend of aerodynamic and acoustic properties.

The mouth channel from the kymograph gave a relatively poor representation of the speech waveform. A much better way of plotting the speech waveform was provided by the Duddell oscillograph. Duddell himself published excellent speech waveforms as early as 1907. An oscillograph of this type was used for the ground-breaking research programme summarised in Fletcher’s classic work Speech and Hearing (1929), which was carried out at Bell Laboratories in the USA, an important research centre which took its modern form in 1925.

But despite its obvious suitability and superiority, the Duddell oscillograph was expensive to buy and run, and also required a lot of ancillary equipment and a photographic darkroom, plus relevant technical skill. The number of laboratories around the world equipped with an oscillograph remained fairly small. For this reason, speech researchers continued to use the kymograph and to look for cheaper ways of plotting the speech waveform.

The primitive version of the speech waveform produced by the kymograph did have some real uses. It showed the timing of major events and the periodicity in strongly voiced sections. There was generally enough information to attempt a segmentation of the wave, so studies of duration were common. Periodicity could be measured in the voiced sections and thus fundamental frequency (voice pitch) could be determined.

The plotting of fundamental frequency this way was so laborious that it was difficult to produce enough material to be of much practical use (for example in language teaching). What was really needed for the study of intonation was an automatic way of tracking fundamental frequency. Two Japanese engineers, Obata and Kobayashi, published the first automatically-obtained fundamental frequency curves of English sentences in 1937. In their device, the input wave is deliberately distorted, reduced to a pulse train, and then the resulting pulses counted (at first by analog means) to produce a time-varying output on a logarithmic scale.

The first international conference on experimental phonetics took place in Hamburg in 1914, just a few weeks before the outbreak of World War 1, and a second conference was held in Bonn in 1930. The modern International Congress of Phonetic Sciences began in 1932 in Amsterdam, and further congresses based on the same format followed in 1935 (London) and 1938 (Ghent). After an interruption occasioned by World War 2, the series resumed in Helsinki in 1961 and established a four-yearly pattern which continues to the present day. The twentieth congress took place in Prague in 2023.

Harmonic analysis of sustained sounds had been possible since the end of the nineteenth century, but real-time analysis of running speech with a bank of filters came in the late 1930s. During WW2 the speech spectrograph was developed at Bell Laboratories, and it was marketed in the postwar period as the Kay Sonagraph. The speech spectrograph became the most widely used instrument in phonetics laboratories in the 1950s and 1960s. The Pattern Payback, invented by Franklin S. Cooper at Haskins Laboratories in the USA, provided a means of synthesizing speech from visible patterns, and much was learned about “acoustic cues” in speech perception. At the same time, formant synthesis was demonstrated by Walter Lawrence in England, and formant synthesizers were developed at various laboratories around the world.

During the 1950s and 1960s many laboratories published their own “working papers” which were largely distributed by exchanges with other labs. Particularly important series of reports came from Bell Laboratories and Haskins Laboratories, and from KTH in Stockholm. Gradually the number of journal outlets for phonetics research increased. Journal of the Acoustical Society of America had always carried speech articles; the journal Language and Speech began in London in 1958, and others followed.

By the 1970s, the field of speech engineering began to emerge, as practical applications of speech synthesis and recognition became possible. A conference on speech processing in Edinburgh in 1987 was the starting point for the Interspeech conference, and the International Speech Communication Association (ISCA) was founded in 1988.

By the mid 1950s, it was understood that digital computers could in principle replace any form of analog hardware for signal processing, though at first the processing might occupy many times real time. By the late 1960s, “laboratory computers” (which were still very large and expensive by present-day standards) began to offer interactive working and graphical displays. Such a computer might be used, for example, to examine and measure the sampled speech waveform, or to control an external hardware speech synthesizer for speech synthesis-by-rule. Desktop computers appeared in the 1980s (the IBM PC was launched in 1981), and speech processing software capable of making spectrograms, extracting fundamental frequency, and performing speech synthesis began to appear over the next decade. Mark Huckvale’s SFS system appeared at UCL in 1987, and Praat (which has become the most widely used speech analysis package) in 1991.

Koerner remarked thirty years ago that the study of the history of phonetics was a relatively undeveloped field, and the same is largely true today. There is currently no single extended work on the broad history of phonetics. The present brief summary may be compared with Heselwood et al. (2013). Otherwise, published work tends to focus in detail on specific periods, laboratories or protagonists.

Excellent accounts of certain aspects are to be found in articles by J. A. Kemp, M. K. C. MacMahon, and H. Tillmann in the Encyclopedia of language and linguistics.

A small international community of researchers form a Special Interest Group on the History of Speech Communication Research (HSCR) sponsored by ISCA. Proceedings of its regular workshops can be found at https://www.isca-archive.org/ (Search: HSCR).

Complete proceedings of ICPhS dating back to 1932 are available at

https://www.coli.uni-saarland.de/groups/BM/phonetics/resources.html#icphs

and many are also at

https://www.internationalphoneticassociation.org/content/icphs

Some Congreses (e.g. Prague in 2023) have included special sessions of papers devoted to the history of phonetics.

Ashby, Michael. 2015. Experimental phonetics at University College London before World War I. In Rüdiger Hoffmann & Jürgen Trouvain (eds.), HSCR 2015: proceedings of the First International Workshop on the History of Speech Communication Research, Dresden, September 4-5, 2015 (Studientexye Zur Sprachkommunikation 79), 118–127. Dresden: TUDpress.

Ashby, Michael. 2016. Experimental phonetics in Britain, 1890-1940. Oxford: Oxford DPhil.

https://ora.ox.ac.uk/objects/uuid:d8bbffae-8a4e-478e-ba65-0f5a5bbd66e1

Ashby, Michael & Patricia Ashby. 2021. Phonetic Teachers and the Reform Movement: evidence from records of the IPA. Language & History. Routledge 64(3). 151–167. https://doi.org/10.1080/17597536.2021.1996085.

Ashby, Michael & John A. Maidment. 2005. Introducing phonetic science. Cambridge: Cambridge University Press.

Ashby, Michael & James McElvenny. 2021. The emergence of phonetics as a field in the nineteenth century. Podcast: https://hiphilangsci.net/2021/03/31/podcast-episode-14/

also published (2022) in James McElvenny (ed.) Interviews in the history of linguistics. Volume I. (History and Philosophy of the Language Sciences 6). Berlin: Language Science Press. DOI: 10.5281/zenodo.7092391, pages 51–60.

Ashby, Michael & Marija Tabain. 2020. Fifty years of JIPA. Journal of the International Phonetic Association. Cambridge University Press 50(3). 445–448. https://doi.org/10.1017/S0025100320000298.

Brown, Keith & Anne H Anderson (eds.). 2006. Encyclopedia of language and linguistics. 2nd ed. Amsterdam: Elsevier.

Collins, Beverley Simeon & Inger M. Mees. 1999. The real Professor Higgins: The life and career of Daniel Jones. Berlin: Mouton de Gruyter.

Fletcher, Harvey. 1929. Speech and hearing. New York: Van Nostrand.

Helmholtz, Hermann Ludwig Ferdinand von. 1885. On the sensations of tone, as a physiological basis for the theory of music. (Trans.) Alexander John Ellis. 2nd edn. London: Longmans, Green & Co.

Heselwood, Barry, Zeki Majeed Hasan & Mark J. Jones. 2013. Historical overview of phonetics. In Mark J. Jones & Rachael-Anne Knight (eds.), Bloomsbury companion to phonetics, 2–20. London: Bloomsbury.

Koerner, E. F. Konrad. 1993. Historiography of phonetics: The state of the art. Journal of the International Phonetic Association 23(1). 1–12. https://doi.org/10.1017/S0025100300004710.

Kohler, Klaus. 1981. Three trends in phonetics: The development of phonetics as a discipline in Germany since the nineteenth-century. In Ronald Eaton Asher & Eugénie J. A Henderson (eds.), Towards a History of Phonetics: In Honour of David Abercrombie, 161–178. Edinburgh: Edinburgh University Press.