Warning: Undefined array key "HTTP_REFERER" in /data02/virt42190/domeenid/www.etag.ee/researchinestonia/rwp/wp-content/themes/rie/partials/components/c-entry-header-post.php on line 9

Why are linguists at the University of Tartu gluing clamps on tongues?


Linguists at the University of Tartu purchased new devices for the phonetics lab. The curiously named “articulograph”  has sensitive sensors that enable the linguists to trace the movement of the tongue and lips while a test subject is speaking.

Its first scientific experiment, which recently won approval from the ethics committee, will be conducted this spring by doctoral student Anton Malmi, who is studying Estonian palatalization.

When pronouncing palatalized consonants, the tongue rises more towards the palate and gives the phoneme a softer i-like sound. In many cases, the palatalized and unpalatalized pronunciation also specifies the meaning of the word, such as with the word palk (meaning log or salary, respectively), hall (meaning gray or frost) or tulp (meaning tulip or column).

In his doctoral thesis, Malmi also analyses the pronunciation of Estonians whose mother tongue is Russian.

“As already known, palatalization in Estonian has been described as weaker than that of Russian since Ariste [linguist Paul Ariste – ed.], but no empirical comparative studies have been conducted so far,” says Malmi, explaining the background of the study.

“Palatalization is an exciting phenomenon,” Malmi continues. “It is usually caused by the vowel ‘i’ following a consonant. But we also use palatalization to pronounce words that in modern Estonian have lost their ‘i’-s in the nominative case. For example, words such as vann (bath) or pann (pan) have to be pronounced with palatalization or they sound very foreign.”

What also makes palatalization so interesting is the fact that in many words its presence or lack thereof determines the meaning of the word. The spelling of these words is identical, but context gives us a clue as to what pronunciation we should use.

The results of the test will also demonstrate how Russian-speaking Estonians living in Estonia form such seemingly similar words and how their pronunciation is affected by the amount of time they have been learning Estonian.

“The test results will expand our knowledge regarding what the acoustic and pronunciation parameters are that distinguish palatalized words from unpalatalized words,” predicted Malmi. The words in which palatalization clarifies the meaning are often important words that are used in everyday conversations and thus require clear definition. For example, when talking about your salary or about tulips, the pronunciation you use plays a vital role.

Malmi intends to conduct a perception test to see whether people using Russian as their mother tongue perceive the acoustic difference between the palatalized and unpalatalized words.

“This was motivated by the fact that when learning a foreign language, we base everything on our native tongue and we hear the world through the filter of our native language. As palatalization in Estonian and in Russian is essentially different, it is expected that in some cases Russian speakers can’t discern Estonian palatalization,” explains Malmi.

Thanks to his doctoral thesis we can find out more about the issues concerning pronunciation of palatalization and this could be useful for those who teach and those who are learning Estonian.

What does EMA do?

According to Malmi’s supervisor, Pärtel Lippus, the electromagnetic articulograph (EMA) is a relatively specific device that can be used for studying the pronunciation of phonemes or articulatory phonetics. To accomplish this, the device tracks the movement of the sensors attached to the tongue and lips in the magnetic field.

“The sensors are glued to the tongue, lips and/or chin of the test subject,” Lippus explains, “after which we can record their movement during the act of speech together with the sound produced.”

There is another similar machine at the Tallinn University of Technology’s (TTÜ) Laboratory of Phonetics and Speech Technology, which is the workplace of Malmi’s second supervisor, Einar Meister, a senior research fellow at TTÜ. It is also the place where Malmi conducts some of his tests.

Anton Malmi, a doctoral student of phonetics at the University of Tartu, is gluing sensors to the tongue of the test subject, Katrin Leppik. Author/source: Pärtel Lippus

The first published study on Estonian language using EMA was conducted at the University of Helsinki and was published last Autumn.

Lippus explains that they studied the pronunciation of the three quantity degrees in the Estonian language.

“Firstly, we discovered that even though we have three degrees of quantity, our pronunciation movements cannot be divided into three. And the second conclusion we drew was that in pronunciation, context plays a significant role.”

What connects the tongue and the eyes?

Within the University of Tartu’s ASTRA project PER ASPERA, funded by the European Union, an eye tracker was acquired alongside the EMA. Linguists will be able to use these for various experiments that study the processing of language.

In the phonetics lab, the eye tracker has been placed into a recording booth. This helps researchers combine tests in such a way that, in addition to tracking the test subject’s eye movements, the test subject will read or answer something out loud and the researchers will record their speech in a noise-free environment.

According to Lippus, eye-tracking tests are quite common in linguistics.

“For example, in December of last year, Kaidi Lõo defended her doctoral thesis in Canada at the University of Alberta. With the help of the eye-tracking test, she studied how the morphological complexity and the occurrence frequency of a word affects reading speed. Part of the study was conducted at the phonetics lab of the University of Tartu but using a portable eye tracker she had brought with her from Canada.”

Eye tracking devices can also be found in other laboratories in Estonia. For example, the University of Tartu’s Laboratory of Experimental Psychology has one of these devices.

“Together with the psychologists, we also conducted a test for analysing the degrees of quantity in Estonian using the eye tracker,” says Lippus.

“We are used to thinking that Estonian is spelled as it is pronounced but there are a lot of words in which the difference of cases is only marked by the degree of quantity that is not marked in the spelling. For example, the genitive case of the word mets (forest) is in the second degree of quantity and its partitive case is in the third degree of quantity.

“The test explored whether reading these words requires more context than when the spelling is unambiguously interpretable, like with the word mõte and mõtte (‘thought’ in the nominative and genitive case, respectively). We are still processing the results, but the initial evaluation shows that the hypothesis was confirmed: More mistakes were made when reading words like mets than with words like mõte.”

The translation of this article from Estonian Public Broadcasting science news portal Novaator was funded by the European Regional Development Fund through Estonian Research Council.

Read more

Get our monthly newsletterBe up-to-date with all the latest news and upcoming events