Speak my language: a first for digital voice communication in Turkmen
People with visual or communication impairments in Turkmenistan will soon be able to navigate the online world thanks to important assistive technology developed in their own language.
Photo: UNICEF Turkmenistan
Muhammetmyrat Orazov, 24 years, from Ashgabat, Turkmenistan uses a screen reader to access information and communicate online.
“In everyday life I use a text-to-speech voice to read news, literature, magazines and communicate with my family and friends. I also read electronic books, learn new things on the computer and use the Internet. But when I download books in Turkmen, or receive text messages and emails in Turkmen, I cannot read it in the Turkmen voice with my screen reader. Instead I have to use other language voices to try to decode the Turkmen text letter-by-letter, which is extremely time consuming and very hard for listening,” says Muhammetmyrat.
A screen reader is software that allows people to navigate the screen on their device, and hear what is presented using a text-to-speech voice. Text-to-speech voices are used to convert text on the screens of smartphones, tablets, computers and other digital devices into lifelike speech. These services, often provided by large multinational companies such as Amazon, Google or Microsoft, are available in multiple languages and often in a variety of voices of different ages, genders and accents. The applications have proven to be hugely beneficial to children and adults with visual or communication impairments. They open up opportunities for education and professional development, and participation in many other realms of life.
However, text-to-speech voices in many under-represented languages, such as Turkmen, are unavailable, making communication and access to information near impossible, or very time consuming, for those who need this assistive technology.
ATscale, the Global Partnership for Assistive Technology, identified an urgent need for text-to-speech voices for under-represented languages, and is making catalytic investments for populations who need better access to these services. Turkmen is the first of four languages targeted in ATscale’s text-to-speech initiative this year; others are Nepali, South Vietnamese and Setswana, spoken in Botswana and South Africa.
In Turkmenistan, ATscale is working with UNICEF, which partners with the Ministry of Education, to develop text-to-speech in the local language Turkmen.
“Assistive technology enables people to live healthier, more productive, independent, and more dignified lives than before and to participate in education, the labour market, and their communities. This text-to-speech project in Turkmenistan provides enormous opportunities to transform lives, benefitting both individuals and society, and ATscale is proud to support this vital investment,” says Pascal Bijleveld, CEO ATscale.
The need for text-to-speech in Turkmen
This project is the start of a new and long-term endeavour to expand access to information and communication in Turkmen. In 2021, the Ministry of Education in Turkmenistan and UNICEF began to discuss the need for a Turkmen text-to-speech voice for screen readers.
Turkmen is spoken by the population at large in Turkmenistan, and in parts of neighbouring Tajikistan, Uzbekistan, Iran, and Afghanistan. But with only about 10 million speakers, Turkmen is not commercially attractive, and with a focus on reaching bigger markets, underrepresented languages like this and others across the world, stay off the radar of major tech companies. This market failure means that the population in Turkmenistan is unable to access essential and potentially life-changing text-to-speech technology.
Thanks to this project, people with screen readers will be able to download the text-to-speech voice in Turkmen. The software will be available free of charge on accessible online platforms. Initially there will be just one voice, with additional voices anticipated for the future.
“Currently reading a 100-page book in Turkmen without a Turkmen screen reader takes 4-5 months for me. With a Turkmen voice I can finish a book in 10 days. The development of a Turkmen voice will enable me to enjoy the books in my native language and make Turkmen language applications and websites accessible to me,” says Muhammetmyrat.
Photo: UNICEF Turkmenistan
Who will use the Turkmen text-to-speech technology?
The Society of the Blind and Deaf of Turkmenistan, with about 3500 members, is the leading public organization bringing together blind and deaf people. However, many people with visual impairments are further estimated to be living in rural areas. The new Turkmen voice technology will be made available to the Library of Turkmenistan for the Blind and Deaf of the Ministry of Culture, which has computer resources and archives, used by many members of the Society of the Blind and Deaf.
People with other types of impairments requiring additional communication support such as speech impairments, and dyslexia will also benefit from this project. The text-to-speech technology will bring voice to the entire operating system where installed, and the voice will also allow localization of apps for special communication needs. Currently, UNICEF is localizing an open source, alternative and augmentative communication (AAC) mobile app, Cboard, to support young children with complex communication needs who otherwise may be left out of early childhood education services.
For children and pupils in the education system, digital accessible books open up new avenues for their learning. For children and adults alike, having access to the Internet, means participation and learning about news, current affairs, social media and engagement with politics, economy and society.
The information technology community in Turkmenistan is also looking forward to integrating the functionality in the applications they are developing.
“The fact that this key functionality for Turkmen is being developed with the support of international experts is great news. I will be able to integrate the Turkmen voice in my application to transform the lives of people with visual impairments in Turkmenistan,” says Suray Yusupova, a young developer from Ashgabat, who is currently developing an app that aims to enhance preparedness of people with visual impairments in emergencies.
Innovation and investment to transform lives
ATscale provided catalytic support to develop open source text-to-speech software in Turkmen in association with UNICEF. LouderPages, a team of software voice developers with visual impairments are working with local collaborators in Turkmenistan. The LouderPages team is distributed across the United States, central and Eastern Europe and most of the members are themselves screen-reader users.
“The international cooperation taking place to carry out a text-to-speech project in Turkmen is a major step for the country in terms of innovation for persons with disabilities. It is hugely significant for implementing the principle of leaving no one behind,” says Nazar Mammedov, Education Specialist, UNICEF in Turkmenistan.
Text-to-speech supports national inclusive education strategy
UNICEF is also partnering with the Ministry of Education in Turkmenistan to bring the technology to pupils across the entire education system, from primary to university levels. The text-to-speech project ties in with the core education priorities of Turkmenistan. The country has a vision and policy in place to develop and sustain inclusive education, with a long-term objective of transitioning from the medical to the social model of disability. This was informed strongly by Turkmenistan’s participation in the United Nations’s Transforming Education Summit in September 2022, which called for education to be elevated to the top of the political agenda. The text-to-speech project is part of the implementation of this broader vision for education in the country. The project also supports the national action plan for child rights, where assistive technologies play a crucial role.
23 May 2023