An uncertain future for African languages
We left the purely utilitarian and "capitalist" perspective distort our perception of the value of African languages. The sad reality is that, due to a lack of immediate economic opportunities, our languages are relegated to the background.
This worrying decline is evident even at the grassroots level, where intergenerational transmission is compromised. Our own grandparents are adopting French and English to communicate with us, thus reversing the natural course of intergenerational transmission. Today, we can speak of a linguistic asymmetry that puts us at a disadvantage. Indeed, a European who comes to our villages does not need to make the effort to learn our language, because they know that we have learned theirs. And yet, when we travel, we are required to take language tests.
Data and Disparities: Inequality in Numbers
The digital age, which was supposed to be one of global connectivity, has created a huge linguistic divide where we see the supremacy of English on the web. A simple comparison between the number of native speakers and the online presence of languages reveals a shocking truth: our African languages are the great forgotten ones of the web.
English accounts for nearly 50% of online content, even though it is the native language of only 4.9% of the world's population. This can be explained by the following factors, among others:
- The technological hegemony of English-speaking countries (United States, United Kingdom, Canada, Australia).
- English is the language of commerce, innovation, and scientific research, which reinforces its digital dominance.
- The network effect: the first web infrastructures were built in English, and the momentum has continued.
In contrast, with nearly 10% of the world's population having an African language as their mother tongue, they represent less than 0.1% of online content. We need to be clear about this: today, a language that is not used in the digital world is a language that is at risk of disappearing in the long term. This disparity can be explained by historical factors (colonial heritage), the lack of infrastructure and internet access, the low production of local content, and the lack of institutional recognition. low production of local content, and little institutional recognition (African governments invest little in promoting local languages in education and technology). This marginalization is therefore not a coincidence, but reflects, among other things, a relentless cultural and economic reality, which is my angle of attack in this document.
Economic Argument: The Nerve Center of Warfare
In a world where everything has a price, languages have become economic tools. Languages with a strong online presence open doors and create opportunities. Our African languages, however, are relegated to the status of folkloric curiosities.
Globalization, as it stands, is a steamroller that standardizes and crushes individual characteristics. Indeed, even from a linguistic point of view, globalization promotes the spread of dominant languages, reinforcing their hegemony.
Although motivations may vary depending on age, social context, and individual goals, statistics from platforms such as Babbel, Duolingo, and UNESCO reports confirm that the top three main motivations for learning languages are economic and migration-related. However, learning African languages is generally linked to identity, which ranks seventh and last on the list of reasons. The fact that identity-related reasons are predominant partly explains the limited scope of African language learning and their perception as "private" communication tools, reserved for family or community use.
African languages and reasons for identity
For many people, learning an African language is a way of reconnecting with their roots, history, and culture. It is a way of reaffirming their identity and preserving a precious heritage. This motivation is particularly strong among African diasporas, where there is a strong desire to reconnect with ancestral languages.
And yes, beyond capitalist reasons, their heritage value is crucial, and that is reason enough.
Issues: Beyond utility
- Identity and heritage: Languages are the repositories of our history, our traditions, and our knowledge. Their disappearance would result in an irreparable loss of our cultural identity.
- Diversity and richness: Linguistic plurality is a source of richness for humanity. Each language conveys a unique worldview. Its disappearance impoverishes our understanding of the human experience.
- Endogenous development: The use of local languages in education, the media, and institutions promotes better knowledge acquisition and more inclusive development.
Mandarin: gaining momentum despite internal challenges
English, French, and Mandarin, to name but a few, have each shaped their countries of origin in different ways. English has consolidated the economic and cultural hegemony of English-speaking countries, along with a certain linguistic imperialism. French, despite its overall decline, still has diplomatic influence. Mandarin is expanding rapidly despite internal structural challenges similar to those faced by African countries, notably internal conflicts related to linguistic standardization, the difficulty of internationalization because it is a language, etc.
The positive impacts of Mandarin's reach are well known, and I can cite, among others:
- Economy & trade: China has imposed Mandarin as the language of business in Asia and Africa thanks to its economic expansion and its New Silk Roads program. In addition, China actively uses Mandarin language learning as a lever of influence through its Confucius Institutes.
- Technological soft power: Unlike English and French, China has built a digital ecosystem entirely in Mandarin (Baidu, Alibaba, WeChat, etc.), which allows it to be less dependent on Western platforms and to control its information flows.
- Internal social cohesion: By imposing Mandarin as the official language and teaching it throughout the country, China has strengthened its national unity and cultural identity.
Can African languages follow the example of China and Mandarin, while naturally taking into account the challenges of our multipolar world, where new technologies, artificial intelligence, and geopolitics are constantly redefining linguistic power relations?
Here is a proposal to strengthen the economic argument by drawing a parallel with the African cultural industry:
The economic paradox of African languages and the cultural industry
If economic arguments justify the dominance of foreign languages, how can we explain the phenomenal success of Afrobeats, Nollywood, and coupé-décalé on digital platforms? These industries, driven by local languages and African cultural expressions, have managed to establish themselves in a competitive world dominated by Western giants.
Take African music, for example: despite the fact that songs are often sung in Yoruba, Lingala, Wolof, or Pidgin, artists such as Burna Boy, Fally Ipupa, and Rema fill stadiums and dominate global charts. Nollywood, with films largely produced in Pidgin and local languages, is now the second largest film industry in the world in terms of volume. These successes prove that cultural production rooted in African languages can not only survive but thrive on an international scale.
This success should make us ask ourselves: why are our languages still absent from digital technology and AI, while our music and cinema are conquering the world? The answer lies in promotion and monetization. If the entertainment industry has been able to capitalize on global demand for authentic African content, it is imperative that we adopt the same approach for our languages. Until they are integrated into digital business models, they will remain marginalized, even in our own territories.
The lesson to be learned? Influence and accessibility are powerful levers. Isn't it true that if we want our languages to survive and thrive, they must generate value, be taught, be used in digital services, and adapt to new forms of communication, just as African music and cinema have done?
Artificial intelligence: hope for our languages
I personally enjoy reading books in foreign languages (German, Turkish) with just one click, thanks to tools such as Google Lens. This is a free tool from Google available in any Chrome browser. Tools such as Google Lens and Google Translate already allow you to instantly translate texts using your smartphone camera, making languages more accessible to users. We are therefore witnessing a paradigm shift : you don't necessarily need to learn a language to be able to understand it. The focus is shifting from memorizing grammar rules to practical communication.
Furthermore, technological developments show that the future of digital technology is increasingly based on voice and speech. Today, more than 50% of Google searches are performed verbally, and the global market for voice assistants, estimated at nearly $30 billion in 2024, continues to grow. With the rise of voice assistants (Alexa, Siri, Google Assistant) and voice interfaces, it is becoming crucial to have massive databases in African languages to enable these tools to understand and interact with native speakers.
Imagine a voice assistant in Yemba, capable of answering a farmer's questions about weather forecasts, giving health advice in Wolof, or helping a shopkeeper manage their stock in Lingala. For these technologies to see the light of day, we need to feed artificial intelligence with as much data as possible in African languages: texts, recorded conversations, translations, etc. However, today, less than 0.1% of online content is in African languages, making our languages invisible to artificial intelligence systems.
What are the relevant areas of training and research?
To avoid this linguistic digital divide, it is essential to invest in research and development, production, and the mass digitization of African languages (see Figure 4), in order to ensure an active presence for our languages in the global digital ecosystem.
In the specific field of AI, the following three areas should be considered.
- NLP (Natural Language Processing): AI enables us to understand and generate human language. Machine translation tools are improving at a dizzying pace.
- NLP (Traitement du langage naturel) : L’IA nous permet de comprendre et de générer du langage humain. Les outils de traduction automatique s’améliorent à une vitesse vertigineuse.
- Text-to-speech: We can now listen to texts in any language, with increasingly natural-sounding voices.
Promising Initiatives, But Still Insufficient
- AI4AfricanLanguages (Leveraging AI for African Low-Resource Languages to Enhance Crises Monitoring) is organizing a workshop symposiumwill address the various stages of text and speech data processing (including collection, preprocessing, and learning) for low-resource sub-Saharan African languages.
- Masakhane : The Masakhane project Masakhane project, a pan-African community of AI researchers, is working on the development of machine translation models for African languages, including Swahili, Hausa, and Yoruba.
- Lelapa AI : The startup Lelapa AI, founded by African researchers, aims to create inclusive language models that take into account the specificities of African languages, with an emphasis on diversity and social impact.
- Vula Mobile : TheVula Mobile app, developed in South Africa, uses AI-based technologies to facilitate access to medical care in rural areas via several African languages. It enables patients and healthcare professionals to communicate more effectively, thereby reducing language barriers in the medical field.
- NTeALan (New Technologies for African Languages). This is a community of young Africans skilled in languages and AI who are working to strengthen the presence of African languages in the digital world.
- Google and Mozilla have also recognized the importance of African languages. Google has integrated translations into local languages in Google Translate, and Mozilla has launched voice collection initiatives to improve speech recognition in Swahili and Luganda.
Similarly, companies such as Orange are collaborating with OpenAI and Meta to integrate African regional languages into AI models, initially targeting Wolof and Pulaar.
Mais il faut aller plus loin, plus vite.
A Call to Action: The Urgency of Preservation
Integrating our languages into AI and NLP is not an option, it is a matter of survival. Languages are not just a utilitarian tool, but in order for them to be useful, they must first be practiced and preserved. Without digital and economic opportunities, African languages are clearly under threat.
- We must promote ambitious language policies.
- We need to digitize our resources, develop technological tools, and create content. By leveraging these technologies, it is clearly possible to develop translation tools, instant communication tools, educational applications, and digital resources in African languages, thereby increasing their usefulness in modern contexts. Knowing that technological tools facilitate their acquisition and use will also encourage non-speakers to learn our languages.
The integration of African languages into AI and NLP technologies is not only feasible, but essential for their preservation and promotion. This will redefine their place in our contemporary societies, giving them greater economic and cultural relevance.
Are we going to let our languages die out amid indifference, or are we going to revive them thanks to the power of digital technology? The choice is ours.
