Linguists help make sense of speech sounds transmitted through space communication radio systems

At the beginning of the 20th century, the rapid development of radio broadcasting and improvement of telephone communication systems led to the emergence of a new branch of acoustics — electroacoustics. In their quest for technological solutions, radio engineers had to overcome many challenges, including acoustic interference associated with speech. 

In order to better understand the sound distortions occurring during radio signal transmission, employees of a Leningrad radio station turned to the University’s scholars. Among the participants of that project were: a prominent linguist Lev Shcherba, the founder of the St Petersburg phonological school, Member of the Russian Academy of Sciences and Professor in the Department of Phonetics and Methods of Teaching Foreign Languages at Leningrad State University; his colleagues — linguists and the University graduates: Sergei Bernstein, the future founder of audio archivistics in Russia; Vera Orfinskaia, the speech pathologist; and other specialists. 

That first collaboration between radio engineers and phoneticians showed the importance of studying speech perception, an aspect that was previously disregarded by linguists.

Back in the 18th century, it was suggested that sound synthesis could help us better understand the physical properties of speech sounds. In 1779, Professor of St Petersburg University, Academician Christian Gottlieb Kratzenstein developed a geometric theory of vowels, thus anticipating Helmholtz’s resonance theory of hearing. When the Academy of Sciences in St Petersburg announced a prize competition for the mechanisms behind the five long vowels A, E, I, O and U in human speech, the researcher constructed a device that could produce these special sounds. In 1780, Professor Kratzenstein won the Academy’s prize for his "vowel organ".

In the late 1930s, the Acoustic Laboratory of the telephone manufacturer "Krasnaya Zarya" (the former Swedish telephone company Ericsson that became the leading manufacturer of communication equipment in the USSR) approached Academician Lev Shcherba with a request. Shcherba was asked to develop test charts of sounds and sound sequences — or, as they were called in telephony then, ‘articulation charts’ — adapted for the Russian language to test telephone equipment. 

Until then, the Esperanto charts, developed by the International Telephone Consultative Committee (ITCC), had been used for these purposes. It was wrongly believed that these data were universal and represented a uniform measurement unit to check the audio quality of recordings in any country. In fact, these charts did not take into account important elements of the phonetic systems of national languages. 

In 1940, Leningrad State University experts Liubov Liapunova and Vera Orfinskaia, under the guidance of Academician Lev Shcherba, developed the first Russian language syllable charts. The charts presented both stressed and unstressed syllables, which were taken from specially provided texts, and took into account the specific properties of Russian pronunciation. Unfortunately, those charts, compiled at the request of telephonists, but without their participation, turned out to be very complicated and of little practical use to them. The efforts to correct and adapt the charts for the practical needs were interrupted by the outbreak of World War II. 

After 1945, research in speech intelligibility on radio received a new impetus. The Laboratory of Experimental Phonetics at Leningrad State University was invited to participate in a collaborative project in conjunction with experts from the Budyonny Military Academy of the Signal Corps (now the Budyonny Military Academy of Communications) and the Research Institute of the Ministry of Communication Equipment Industry of the USSR. They worked together on improving the channels through which speech messages are transmitted and investigated speech perception in challenging conditions with noise interference. That research programme was instrumental in the development of space communications.

The collaborative research programme with the Military Academy of Communications was classified. Only a few people were aware of the project owner, and the data obtained were published in closed access scientific journals.

The projects of the Laboratory of Experimental Phonetics at Leningrad State University were headed by Professor Lev Zinder (1903-1995), later joined by Professor Alla Stern (1942-1997), a renowned expert in speech psychology and psycholinguistics. They were the scholars who developed the test charts used to assess the quality of Yuri Gagarin’s microphone during his space flight. 

In the 1960s-1970s, young doctoral students: Nina Liubimova; Liudmila Verbitskaya; Lia Bondarko; and Natalia Sviatozarova — joined the research programme. They were scholars of the St Petersburg phonological school founded by Lev Shcherba. Each of them studied a different aspect of acoustic phonetics. For example, Lia Bondarko collaborated with the Pavlov Institute of Physiology as a research engineer. Her main focus area was the syllable. Based on acoustic data, she set up a series of experiments using a multi-sensor device to examine the contributions of all articulators. When a speaker produced an utterance, the device measured and recorded the work of the entire articulatory system. 

According to the recollections of Nina Liubimova, Professor in the Department of Russian as a Foreign Language and Methods of Its Teaching at St Petersburg University, the researchers working on the project were "incredibly decent and friendly people, loving their work immensely".

When we conducted that research, we did not realise where the research findings would be implemented. Everything was classified. We described acoustic signatures of the recordings we had been given from a linguistic perspective. For instance, military radio engineers did not use the word "sound", they only used the word "letter". When I had to present the research results to big scholars with military insignia, I was shaking like a leaf. I printed out my materials and arranged them in sets in advance. I presented the research findings and explained them in detail. The generals even praised me then. If we managed to make at least some notes on the margins of science, it is already an achievement to be proud of.

Nina Liubimova, Professor in the Department of Russian as a Foreign Language and Methods of Its Teaching at St Petersburg University

The research findings of the Laboratory of Experimental Phonetics provided practical insights for language teaching methodology. The linguists thoroughly studied how to work on articulation and how to teach speech sounds. The research also had important theoretical implications for linguistics as a science — the language was described quantitatively — on the basis of objective data. 

The Laboratory began to study formant structure of stressed vowels and the duration of sounds in individual words and in connected speech. For example, if we compare the duration of vowels in individually pronounced words and in connected speech, it will be evident that the stressed vowels in a phrase are always shorter than in individual words. Unstressed vowels are not always phonologically reduced in connected speech, and only insignificantly in case they are reduced. 

Another important feature of Russian phonology was revealed during that collaboration between phoneticians and radio engineers regarding the nature of the Russian system of stress, readily observable in speech synthesis. The researchers were able to confirm the long-standing experimental phonetic data indicating that intensity, unlike duration, is not an essential feature of the stressed vowel in Russian. 

That research programme bore fruit for both radio engineers and linguists. For example, the compilation of phonemic charts required statistical data on the structure of the Russian language. Such data had not been collected before, including: the frequency of occurrence of individual phonemes; different combinations of consonants; and different types of syllables. Radio engineers learned that in linguistics there is a unit of sound, called a phoneme, which should be the subject of acoustic research as well. 

The research programme’s findings helped the radio engineers in compiling test charts, which were later used to assess the quality of the microphone used by Yuri Gagarin during his space flight. Thanks to that research, the engineers adjusted the radio signal and microphone so that they were able to transmit directly from space the famous: "The flight is proceeding normally".