Originally published on Global Voices

Photo of some of the kidnapped victims rescued by the Nigerian army. Photo by the Nigerian Army, via Facebook. Used with permission.

Tragedy struck some communities in Kwara State, Nigeria, on June 30th, when armed bandits ambushed people in Gada Woro community, Edu Local Government Area (LGA), resulting in the death of three vigilantes and leaving several security operatives injured. Seven days later, three people were kidnapped in the neighboring Gbugbu community.

In another horrible incident on September 22, 2025, bandits launched a deadly attack on the Oke Ode community in Ifelodun LGA, Kwara State, Nigeria. They killed 11 forest guards and the community head, and also kidnapped some of the residents. 

These attacks came amid a rising wave of attacks, kidnappings, and killings that have gripped communities in Kwara South and Kwara North senatorial districts in recent months.

A day after the deadly attack at Oke Ode community, September 23, bandits attacked a mining site at Agbonran community of Ifelodun LGA, killing a police officer connected with the company and abducting three other policemen. Four days later, bandits attacked the Tsaragi community of Edu LGA. After firing multiple gunshots into the air, the armed gunmen abducted a businessman named Suleiman Ndana. 

The armed bandit groups, seeing no government response or reprisal, stepped up their attacks, moving from small-scale abductions to attacking larger communities. Before the Oke Ode killings, the bandits had attacked communities such as Babanla, Baba Sango, Oreke, Ganmu Aliheri, and many others. Human Rights Watch (HRW) estimates that at least 638 villages have been attacked over the last two years across Nigeria.

These attacks have led to at least 10,000 deaths over the last two years and the displacement of many residents of the affected communities. Due to fear of potential attacks, many have fled their homes and relocated to Ilorin and other neighboring communities to seek refuge. 

The Kwara State government has closed 45 public schools in the affected areas. Residents of the neighboring communities are living in fear and appealing to the government and security agencies to intervene and protect their lives and properties. 

Reactions to the killings

After receiving intelligence about a potential attack on their community, 15 local vigilantes in Isanlu Isin attempted to repel suspected bandits in the area on September 14th. The encounter with the bandits was bloody; the bandits killed two of the local vigilantes, abducted six others, and burned 27 motorcycles to ashes. 

To draw the government's attention, angry residents in the Isin community blocked the Ilorin–Lokoja highway on September 12th. In October, young people in Tsaragi town in Edu LGA took to the streets to protest against kidnapping and banditry in the community. The Kwara South Youth Congress (KSYC) had also called on the President of Nigeria, Bola Ahmed Tinubu, to declare a state of emergency in the state. 

Citizens have also taken to social media with hashtags such as #KwaraIsBleeding and #SaveKwara to pressure the government to take action. 

While the national government has largely been silent, the Kwara State government has tried to take measures to address the issue. The state Governor, Alhaji Abdulrahman Abdulrazaq, visited the affected communities. He condemned the attacks, mourned the victims, and requested the deployment of more federal troops to curb insecurity in the state. He assured the residents that his administration will do all it could to protect their lives.

 After attending a security council meeting on October 2nd, 2025, he said:

We will fight insecurity with every resource at our disposal. Kwara will not be a hiding place for criminals. To any criminal who seeks to destabilize the peace in Kwara, you may run, but you cannot escape justice. Kwara belongs to its people, and their peace will be defended.

Other stakeholders have also taken action to address the insecurity. To enhance security measures in response to the escalating insecurity in the region, the Nigerian Army Headquarters directed the General Officer Commanding (GOC), 2nd Division in Ibadan, to relocate to Kwara State on September 28th, 2025. 

The Nigerian Senate also asked the Federal Government of Nigeria to establish a permanent military base near the forest belt of Ifelodun LGA in Kwara South as a quick reaction force to counter armed attacks in the area.

Operation Fasan Yanma

The Nigerian Army personnel have been working to restore peace in Kwara State. The Nigerian Army troops of 2 Division/Sector 3 Operation FASAN YANMA successfully rescued 21 kidnapped victims that included 4 Chinese nationals. Nine more kidnapped victims were rescued by the troops of Operation Fasan Yanma on October 13th, 2025. 

The Commander of the 22 Armoured Brigade, Brigadier General Nicholas Rume, handed over the 23 rescued hostages to the Governor of Kwara State on October 28, 2025. While receiving the hostages from the commander of 22 Armoured Brigade, the Governor commended the Nigerian Army and announced plans to deploy forest guards after the army’s cleanup to prevent ungoverned spaces and allow farmers to return to their farmlands. 

The cause of the escalating violence

The destabilization of Kwara State and North-central Nigeria has been linked to different factors such as clashes between farmers and herders, the emergence of new so-called terrorist groups, governance gaps, and bandit spillover.

Some stakeholders have claimed violent Fulani herdsmen are behind the attacks, while others blame bandit spillover due to counterterrorist campaigns in the north. For example, due to intensified counterinsurgency operations in northeast and northwest Nigeria, many attackers are fleeing into the forests of Kwara State and its neighboring areas to seek refuge. According to a report by Daily Nigerian, the Governor of Nasarawa State, Abdullahi Sule, claimed the Wulowulo and Lakurawa group, splinter groups of the Boko Haram sect, an internationally recognized terror group, are targeting farmers and villagers in North-central Nigeria, particularly in Kwara State.  

In a recent Facebook post, Sanusi Lafiagi, a public commentator and lecturer in the Department of Islamic Studies at Al-Hikmah University, expressed optimism about the improved security situation in Edu and Patigi Local Government Areas in Kwara State. Lafiagi wrote on his Facebook page:

Alhamdulillāh it's been a couple of weeks since the last reported cases of banditry in Edu and Patigi Local Government Areas. Peace and quiet are gradually returning to the once troubled and deserted villages. Kudos to the Armed forces for their sacrifices and their relentless and sustained battle against these criminal elements. Also, Kudos to the government at all levels for fulfilling their constitutional duties in spite of the challenges and setbacks.

Lafiagi suggested some solutions to the authorities to continue improving the situation. He wrote:

However, our appeals to the authorities are as follow:

1. 1. Continue the combing of our forests to destroy and eliminate the criminals completely;
   2. Ensure the immediate and safe return of all hostages to their homes and families;
   3. Establish a military base in Kwara North similar to the one in the South;
   4. Make a meaningful use of the forests, lest it becomes a base and safe haven for criminals;
   5. Equip and support the local vigilantes with sophisticated weapons to complement the efforts of the military;
   6. Support the victims of these violent crimes financially and psychologically to overcome these challenges;
   7. Improve basic amenities such as road, hospitals, schools, electricity, etc., in these communities to ease livelihoods;
   8. Establish and finance local intelligence gathering to build trust and mutual relationship between the authorities and the locals;
   9. Reduce unemployment and consumption of illicit drugs among the youth to be able to eradicate crime;
   10. Monitor and regularise mining activities, as a lot of these crimes are allegedly linked to mining;
   11. Increase border patrol and security. The porosity of our borders is a major challenge to our security.

May Allāh protect us all.

At the moment, the authorities are working to restore calm and order. Nigerian Armed Forces are still combing the forest belts in Kwara State and neighboring states such as Kogi State and Niger State in order to flush out criminals hiding in the forest. Many residents are hoping that peace and stability will return to the affected areas soon.

Originally published on Global Voices

Image credit Roshan Shreshtha.

There are nearly 4,900 registered print media outlets in Nepal. Around eighteen news outlets operate exclusively online through digital platforms such as Facebook and YouTube. By law, online media must register with Nepal’s Department of Information and Broadcasting; however, not all outlets currently comply with this requirement.

Standing out among these registered outlets is Khoj Samachar, an independent digital media platform that raises awareness among young people, challenges corruption, and amplifies underrepresented voices in Nepal. Its founder, Roshan Shrestha, is an investigative journalist from Nepal. His approach relies on digital platforms such as Facebook, YouTube, and a dedicated app to draw public attention to marginalized stories and pressure authorities to take action. The platform publishes content in Nepali, in both text and video formats.

Abdulrosheed Fadipe of Global Voices interviewed Roshan via WhatsApp to discuss his journey as an investigative journalist and how he uses digital platforms to promote social change, strengthen good governance, and encourage citizen engagement in Nepal.

Global Voices (GV): What inspired you to start Khoj Samachar, and what challenges did you face in building it from scratch?

Roshan Shrestha (RS): I started my journalism journey during my school years at Shree Bachchhala Devi Higher Secondary School, Nagpuje, Sindhupalchok District, in Bagmati Province of Nepal. When I was in class 8, a man who was the presenter at Avenues Television came to my school to teach students about journalism. That training programme planted the seed of journalism in my mind.

My interest in journalism grew after a devastating earthquake occurred in Nepal in 2015. Although the epicenter of the earthquake was located in Gorkha district, many houses in my village, which is located near the Bhotekoshi River, were destroyed. Many people lost their lives. The Tataponi border crossing with China was also closed. I had just completed class 11 then. I was using a simple Android phone because I was from a low-middle-class family that depended on agriculture for a living. With that phone, I started filming and uploading videos of the earthquake's destruction of villages on a YouTube channel. This was how my journalism career began.

In 2021, I officially launched Khoj Samachar. I did not receive any financial support when I started nor did I have a big team. Now we have a team of four individuals broadcasting news in both English and Nepali.

GV: How do you maintain editorial independence and ensure your outlet's content remains unbiased and free from external influences?

Roshan Shrestha (RS): We do not take money, favors, or support from any political party, business group, or outside donor. From day one, Khoj Samachar has survived only through the income we earn from Facebook and YouTube ads. With those limited earnings, we pay our team and run the entire platform. This is not always easy, but it gives us full freedom. Nobody can tell us what to say or what to hide. We can question leaders, expose corruption, and speak for the people without fear of losing sponsors or upsetting funders. Our only loyalty is to our readers and viewers. For us, independence is more valuable than comfort. Even if the road is difficult, we choose truth over compromise — that is why our content remains unbiased and trusted.

GV: How do you prioritize stories that amplify underrepresented media voices, and what impact do you hope these stories have on your audience?

Roshan Shrestha (RS): At Khoj Samachar, we always give priority to the voices that big media often ignores — farmers, workers, rural youth, and ordinary citizens who face injustice. Many times, people from villages directly contact us to share their pain, because they feel unheard elsewhere.

We have highlighted stories like that of a person struggling to get a bridge built in his village, a road left unfinished for 10 years, or a family trapped in hardship without support. By showing these issues through videos, we try to bring public attention and pressure the authorities to act.

Our hope is that by giving a platform to these underrepresented voices, young people will feel more connected, more aware, and inspired to raise their own voices. We want our audience to see that journalism is not only about leaders and cities, but also about the real struggles of common people — and that truth can bring real change.

GV: What strategies do you use to reach audiences beyond capital cities, and how do you measure the impact of your reporting in these areas?

Roshan Shrestha (RS): From the beginning, our focus was not only on Kathmandu but also on rural and remote areas. We create stories that reflect the real problems of villages — like missing bridges in Humla, neglected roads in Dailekh, or migrant workers’ struggles. Many of these issues are sent to us directly by local people through phone calls, Facebook messages, or our app. By turning their voices into reports and videos, we make sure their problems are heard nationally.

To measure impact, we look at the response from citizens and authorities. For example, in Saudi Arabia, when stranded workers contacted us, our coverage led to parliamentary debates and direct government action. Similarly, when we cover local problems, we monitor community feedback in comments, shares, and messages, and also check if the government or local representatives respond. For us, the real impact is when ordinary people feel their voices are finally heard and some positive change begins.

GV: What are some of the biggest challenges you've faced as an independent journalist, and how have you overcome them?

RS: There have been many difficult moments in my journey, but the most painful ones came when some people, who are unable to accept the truth, started abusing my family and even threatened physical attacks in public. As a journalist, I can tolerate criticism and risk, but when my family is dragged into controversy for my work, it is truly unbearable.

Speaking the truth, especially against corruption and powerful figures, always brings political pressure. Many tried to silence us or scare us into stopping, but journalism loses its meaning if we compromise with fear. What keeps me going is the trust and support of the public and the inner courage to always speak the truth. The emotional burden is heavy, but I continue because ordinary people rely on us to be their voice, and that responsibility gives me the strength to move forward.

GV: How do you envision the future of Khoj Samachar, and what steps are you taking to ensure its sustainability and continued impact?

RS: My main goal is to build Khoj Samachar as a platform based only on truth and facts, not dependent on external funding or political power. The message I want to pass to youth is that they can succeed in journalism by standing for the truth, even if they are alone. I am working to add more features to the Roshan Shrestha App so that it will be useful for the public. This step will help us sustain our work, expand our impact, and keep serving society with independent journalism.

GV: Why did you name the Roshan Shrestha App after yourself?

RS: By putting my own name on the app, I wanted to assure users that this platform is genuine, transparent, and trustworthy. It is not for self-promotion but rather a matter of responsibility: if my name is attached, people know they can rely on it.

Originally published on Global Voices

A group photo of members of Tyap Wikimedia User Group. Image by Friday musa on Wikimedia Commons (CC BY-SA 4.0 Deed).

Tyap is a dialect cluster of Plateau languages that is spoken in the Middle-Belt region in Nigeria. Like many of the languages spoken in this area, Tyap language is currently affected by Hausa language dominance. However, there is still hope as a group of language advocates are using a Tyap Wikimedia platform launched in 2020 to promote the language.

In a chat with Global Voices, Kambai Akau, language advocate and Executive Director of Tyap Wikimedians Organization, discussed the ways they are trying to revitalize and preserve the Tyap language.

Global Voices: Please tell us about yourself

KA: I am a minoritized language advocate. It is my desire that minoritized languages are preserved from extinction because languages are identities of people. I am from Southern Kaduna, a part of Nigeria where numerous tribes live and many Niger-Congo and Chadic-Afro-Asiatic languages are spoken. Hausa is majorly spoken in Northern Nigeria, and it is a threat to the existence of other languages. I majorly work on my mother tongue, the Tyap language and other Platoid languages. 

GV: Can you share some background information about Tyap and its current status

KA: Tyap language is not taught in schools. At the moment, the language has a few literary materials which are primers written in the early ’90's and a few unpublished materials. Currently, there is a language development board called Tyap Language Development Board that is made up of elderly people without much financial resources to fund this project.

There are a number of materials including the A̱gban swáng a̱lyiat (the Tyap dictionary) project which we hope to be published this year through our partnership with Wikitongues. Tyap also has Wikipedia and Wiktionary pages, but those who are literate in the language are few. Because of this high level of illiteracy, many of those who can speak the language do not know how to write it. People living in the villages prefer speaking English to their children instead of Tyap. This did not used to be the case fifteen years ago.

Kids are growing up and viewing Tyap as inferior to English and Hausa. This is a pathetic situation, and if we do not make efforts to save this language, it is going to die within the next twenty-five years.

GV: What motivated you to start preserving the Tyap language through Wikipedia?

KA: I grew up as a child who didn't know how to speak my native language. So, I always felt ashamed of the language itself. But as I became a teenager, I started becoming aware of my identity because I grew up in Benin City and was surrounded by classmates who bore Indigenous names and spoke their languages, while I was the only one who did not speak my Indigenous language and bear foreign names.

Later on, I read a lot of history books, through which I learned about anthropology and different cultures. Something happened around 2007; I listened to a lecture delivered by Dr. Charles Soludo, the then Governor of the Central Bank of Nigeria. In the lecture, he told the story of how he readopted his African name. This story also changed my views about my identity. I tried to translate some English words into Tyap. When I visited my village later, I asked some of the elders to tell me the equivalent of some English words in Tyap. I discovered that not all of them could speak the language; some weren't even interested in it. That's when I realized the language was dying. If the elders couldn't speak the language fluently as of 2007, imagine the current state of the language. All these factors motivated me to take action.

GV: How do you ensure the accuracy and authenticity of the Tyap language content on Wikipedia, given the lack of standardized orthography or documentation?

KA: To ensure the accuracy and authenticity of the Tyap language content, the Tyap dialect used in all Wikimedia projects is the Mabarado dialect (proper dialect). Contributors who speak other dialects are welcome to use their own dialect, and we would translate it back to the Tyap proper dialect. We call it the Tyap proper to differentiate between the Tyap variety spoken in Atyapland and other variants, such as Gworok and Fantswam, which their speakers consider distinct languages. On the Tyap Wikipedia, an administrator reviews texts and edits grammatical errors provided by contributors. A New Testament Bible is also available in Tyap, and we use its spelling as the standard, with some exceptions, such as using ‘c’ instead of ‘ch’ to represent the /tʃ/ sound.

The Tyap Wikipedia and Tyap Wiktionary features a modification tool that enables speakers of other Tyap dialects to translate articles into their respective dialects. Although the other variants lack their own orthography, we are working on creating a unified orthography for these dialects.

Photo of participants at a Tyap Wiki training workshop. Image by Friday musa on Wikimedia Commons (CC BY 1.0 Deed).

GV: What role do you think Wikipedia can play in promoting language preservation and revitalization efforts for endangered languages like Tyap?

KA: Wikipedia is doing a great job in preserving endangered languages by providing a free platform for the documentation of languages like Tyap. Some people might not be able to create or maintain a website for documenting this language if they were to pay for the domains. The foundation can do more by supporting the publication of tangible materials such as books in minority languages. They should consider the economic disadvantages faced by the speakers of some languages who do not have access to the internet or mobile phones. 

I am thinking about the possibility for the foundation to establish a printing press similar to that of the University presses such as Oxford University Press.  This printing press could print the materials created by volunteers at zero cost and distribute them worldwide so that people can easily access them.

GV: Can you share any success stories or notable achievements in your efforts to preserve the Tyap language through Wikipedia?

KA: The creation of the Tyap Wikipedia project is a success story itself. It became the seventh language-specific Wikipedia to be approved in Nigeria in just two years. That's a great achievement in a country where over 500 languages are spoken.

GV: What are some of the most significant challenges you face in terms of resources, funding, or support for your Wikipedia-based language preservation efforts?

I personally struggled with internet access due to faulty smartphone, and an unstable internet connection and power supply in my area. When we started the Tyap Wikipedia project we faced difficulties while translating technical terms. We had to invite contributors to help coin new terms, which were submitted to the Tyap Language Development Board for review. Unfortunately, the board was not cooperative. We have a limited human resource capacity. This is slowing down the work.

In terms of funding, we continue to fund some activities from our pockets. After the programs, we struggle to fund other activities. Wikipedia provided a general support fund for us last year, but the fund was not enough. With this support, we were able to provide mobile devices and internet subscriptions for some contributors. We hope to do more in the future.

GV: How do you envision the future of the Tyap language?

KA: I envision a language that has sufficient literary materials online and hard copies that people can access wherever they are in the world. I dream about a future where children grow up speaking Tyap because their parents speak the language to them during childhood. I envision a future where Tyap language will be taught at all levels of education.

Originally published on Global Voices

Kerry Kan (Assistant producer) capturing Suvan Pal (documentary producer) interviewing Professor Chien Yuechen. Photo provided by Mr. Suvam Pal. Used with permission.

Taiwan is one of the most linguistically diverse countries in the world. A total of 26 languages are spoken on the island, which includes Mandarin Chinese, the official language, Cantonese, English, Japanese, and 22 Indigenous languages. Fourteen of these indigenous languages are endangered. 

Yilan Creole is one of the 14 endangered Indigenous languages in Taiwan. The language is a contact language, a language that emerges or is used as a means of communication between groups with different native languages. It is often used for trade, commerce, or other interactions. In this case, Yilan is a Japanese-based creole language developed from colonial interactions between the speakers of the Indigenous Taiwanese Atayal language, Seediq language, and Japanese colonists during the period when Taiwan was under Japanese colonization (1895–1945). It is now mainly spoken by a small number of elders across four villages on the east coast of Taiwan. It still lacks official government recognition and is not taught in schools.

However, there are ongoing local efforts to keep this endangered language alive. One recent effort is a new documentary on the speakers of the language produced by TaiwanPlus, a Taiwan-based independent news outlet. 

Speaking to Global Voices via WhatsApp, Suvam Pal, a seasoned international journalist at TaiwanPlus, shared the story behind the documentary.

The documentary crew with Professor Chien and Tungyueh villagers. Photo provided by Suvam Pal. Used with permission.

Global Voices: Please tell us about yourself.

Suvam Pal: I'm a cross-platform international journalist with over 20 years of experience in five of the world's most ethnically diverse and culturally vibrant countries: China, India, South Africa, the UAE, and Taiwan. I have been working for TaiwanPlus, a state-funded media in Taiwan, since two years ago. Before working in Taiwan, I worked as a journalist in India, and then I worked for a TV station in South Africa. I also worked for a state-controlled media, CGTN, for nearly five years in Beijing, China. In between, I also worked as a documentary filmmaker in Dubai, United Arab Emirates. I have published several acclaimed books through prestigious international publishers like HarperCollins and Routledge.

GV: Can you share some background information about Yilan Creole and its current status?

SP: Yilan Creole was discovered in 2006 by Professor Chien Yuechen, a senior professor at the Indigenous Studies Department of National Dong Hwa University, Eastern Taiwan, Hualien County. She discovered this language along with Professor Sanada Shinji, a Japanese professor at the University of Osaka, whom she was working under as a research scholar.

During their research, they realized the language met all the requirements for a language to be designated as a creole language. This happened because Taiwan was a colony of Japan for 50 years, between 1895 and 1945 (the end of World War II). Japanese was the de facto official language in Taiwan then. A lot of older generations still speak a bit of Japanese till today. Some of them can speak, read and write Japanese. A lot of people heard people speaking a form of Japanese in Taiwan, but nobody realized that it was a creole language.

Professor Chien found out that the language is a creole spoken in four villages in Yilan County. I went to two of the villages to shoot the documentary. The distance between Aohua to Tungyueh is around 45 minutes to 1 hour drive. The residents of these four villages called the language local names because they used to think it was a local dialect. Professor Chien studied the language and realized that it matched all the criteria to be called a Creole language. Since the language was discovered in Yilan County, she named it Yilan Creole. Recently, Professor Chien did research for the Indigenous languages Department of the Taiwan government and she discovered that there are less than a thousand fluent speakers of the language at the moment.

The language is the only known Japanese-based Creole in the world. It is a mixture of Japanese, and the language of two Indigenous tribes in Taiwan, Seediq and Atayal. The language of these two tribes mixed with Japanese and that's how Yilan Creole was created.

GV: What inspired you to create a documentary about the language?

SP: I would like to inform you that I'm the first ever Indian journalist to work in Taiwan. My work in Taiwan has made me realize that the stories about Taiwan are less told in the world.

When I came here, I found two different divisions in media coverage here. The first one is the Taiwanese local media that has a different coverage style. The local journalists see Taiwan from a different perspective because they are the locals. And there is the Western media whose reporters have a different perspective.

I noticed that there is a gap in the way both sides report their stories and I want to bridge the gap. The Western media are always focused on geopolitics, defence and trade. The local journalist generally covered local politics and other stories. So I started covering stories that are not covered by both sides.

That's how I got to know about Yilan Creole. I was surprised that a story about the language has not been told in English even though a Japanese media organization produced a documentary on the language in Japanese many years ago. Now it is important to tell the story because there was a recent research on the language that was initiated by the government. The language was removed from the list of recognized languages in Taiwan. That was how I got the kick to tell the story. Another thing that inspired me to create the documentary is my love for languages. I am from India, a linguistically-diverse country, and I speak three languages. I have always been fascinated by languages, which is part of the reasons I came to work in Taiwan. 

GV: How did you work with the language community to ensure their stories and perspectives were accurately represented?

SP: As a journalist, fact-checking is a very important part of my work. I did a lot of verification. I spoke to experts and linguists because I think understanding the culture and socio-cultural nuances are important factors a journalist must consider while covering a story. I also did thorough research and asked the professor to crosscheck the first draft of the work. She also got many Yilan Creole native speakers to crosscheck the work to ensure their perspectives were accurately represented.

GV: What impact do you hope this documentary will have on the language community and broader audience?

SP: Ninety-nine percent of the Taiwanese I have shared and discussed this documentary with told me that they did not know about the existence of Yilan Creole prior to our discussion. These include some of my colleagues, university professors, journalists, editors, researchers, etc. This documentary has made them aware that there is a need for the revitalization of the language. I have seen the number of views the video has on our YouTube channel [over 83,000] and how people are talking about it in the comment section. This gives me a renewed hope that the language can be saved. Another significant impact is this interview because I believe a broader audience outside Taiwan will become aware of the language.

Originally published on Global Voices

Aremu Adeola presenting a paper at the AfricaNLP workshop co-located at the International Conference on Learning Representation (ICLR) in Vienna, 2024. Image used with permission.

As more and more key aspects of daily life migrate online, language inclusion can be a key component of ensuring equal access for all in digital spaces. However, many African languages lack the resources to develop language technologies and fully migrate online. This typically does not affect the most spoken dialects of these languages, which are often used as the standard dialect, but often affect less common language dialects.

Most efforts to create resources for these low-resourced languages are concentrated on the standard dialects, while many regional dialects that are spoken by millions are neglected. 

The Yoruba language is spoken by 47 million speakers in the world. It is mostly spoken in  Nigeria, Benin, and Togo, with smaller migrated communities in Côte d'Ivoire, Sierra Leone and The Gambia. Although the standard dialect of this language has received considerable attention from Natural Language Processing (NLP) researchers, no resources have been developed for its non-standard dialects. To address this problem, a group of linguists have developed YORULECT, a high-quality, contemporary Yoruba speech and text data parallel corpus for four Yoruba regional dialects. 

Speaking to Global Voices via WhatsApp, Aremu Anuoluwapo, a computational linguist who is currently pursuing a master's degree in computational modelling of languages and cognition at the University of Trento, Italy, shared the motivation behind this project. 

Global Voices (GV): Can you tell us a bit about your background and what led you into the field of computational linguistics? 

Aremu Anuoluwapo (AA): I am a linguist by training. I studied Linguistics and African Studies at the University of Lagos. I was introduced to computational linguistics by a mentor, Kola Tunbosun, during my undergraduate years in the university. Then, I worked on some projects that are related to data collection, cleaning and annotation projects. I started gaining experience and collaborating with professionals from multinational companies such as Google, Microsoft, etc, in my third year in the university. My interest in using computational tools to analyse, predict, or transform languages grew from there. 

GV: Can you tell us what motivated the creation of YORULECT? 

AA: Oreva Ahia, my colleague who is a PhD student in Computer Science at the University of Washington, United States, told me about an idea she has on dialectology. This reminded me of a course I took, on Yoruba and its dialectology, during the third year of my undergraduate studies. We learnt about some scholars who have done some works on dialects such as Ẹ̀gbá, Èkó, Ọ̀yọ́, etc, and how the standard Yoruba is primarily drawn from the Ọ̀yọ́ dialect. I found the course interesting and had always wanted to do something on dialectology.

From studying that course, I realised that the word for ‘stool’ differs between the dialect spoken in my hometown in Yorubaland and standard Yoruba. There are other communities that also have distinctive dialectological names for several items. I was curious about this.

Later, while attending a conference in Spain, I travelled to Paris to discuss the idea with Oreva. We designed the framework to execute the project. When I returned to Nigeria, I travelled to specific communities where the dialects we decided to work on are spoken. Deciding on the dialects to work on was a bit technical because there is a division in Yoruba dialectology. There is Southwestern Yoruba, Southeastearn Yoruba, Northeast Yoruba dialects, etc. We wanted to touch all these dialectological divisions. 

One of the reasons we decided to do this project is because of the growing application of AI and machine learning in the tools we use today. We wanted to ensure that low-resourced dialects of low-resourced languages are also represented in technology. 

GV: Could you describe the specific dialects you are working with and explain why these were chosen as a focus?  What are some of their unique linguistic features that pose challenges for NLP systems? 

AA: The dialect we worked on are Ìjẹ̀bú, Ifè, Ilaje and Standard Yoruba. We chose these dialects because they belong to different dialectological divisions of the Yoruba language. We also wanted to have a good representation of these languages in technology. Another reason is because we wanted to do a comparative analysis of how existing NLP systems understand the dialects of the languages before finetuning them. We tested it on Automatic Speech Recognition (ASR) and Machine Translation (MT) and the performance was bad. We also did some finetuning to augment the performance before it became a bit better. 

Some of the linguistic peculiarities we discovered is that there are some letters existing in the dialects that are not in Standard Yoruba alphabets. The Ilaje dialect is a good example of dialect with some different letters and sentence construction. The language structure is similar across all the dialects, although there are some different syntactic arrangements. Our findings show that Ifè dialect has the highest degree of similarity with Standard Yoruba, while Ilaje has the lowest degree of similarity with Standard Yorùbá. We are planning to do more work to expand the research. 

GV: Many African languages are primarily spoken.  How do you approach the challenge of collecting and curating language data for Yoruba dialects that have limited written resources or standardized orthography?  

AA: This was a tough challenge for us. Some dialects do not still have a big population of people who can write those dialects. We were able to surmount the challenges because some of these dialects have scholars who have worked on them.

I always try to tell linguists who want to do this kind of NLP project to collect the speech data first and recruit the native speakers to transcribe them. Doing this will help you get the raw form of the language. It will help you see the many phonological processes that exist in the language. 

GV: What methods do you use to ensure data quality and authenticity? 

AA: We work with the native speakers to collect the data. For the sake of data quality and authenticity, we recruited human evaluators, who are also native speakers, to rate the performance of the ASR systems by giving feedback on the accuracy and quality of their transcription. 

GV: What other challenges did you face when developing YORULECT?

AA: Training the models was challenging. We had to fine-tune them. The linguistic distinctiveness poses a major challenge because the models have not been previously exposed to that kind of data. Some of the dialects did well, while others didn't. This could be because of syntactic arrangement and letter representation.

GV: What are your long-term goals for this work?  

AA: The long-term goal is to strike a new direction in low-resource language research. When the NLP community is discussing low-resourced languages, it is usually about the standard dialect of the languages. They do not consider other dialects. As long as these dialects are still spoken, why not also build tools for the communities that are speaking them? Once the conversation starts people start doing dialectology research on specific dialects of language

Originally published on Global Voices

Yoruba traditional religion adherents during ÈṢÙ is not Satan walk in Osogbo. Photo by Iyanifa Egbetunmise, used with permission.

On December 24, Yoruba traditional religion adherents took to the streets in Lagos, Osogbo, Nigeria and other cities across the world to protest against a historical error that has been causing religious discrimination for centuries. The traditionalists were seen on the streets, donning vests and holding placards and banners with inscriptions. For centuries, the adherents of Abrahamic religions have equated a Yoruba deity called Esu (also called Eshu) with Satan, who is a figure regarded as the embodiment of evil and an adversary of God in Islam, Christianity, and Judaism. The Yoruba are a prominent people group in West Africa, with large populations in Nigeria, Benin, and Togo, and a diaspora around the world. Many Yoruba people practice the Yoruba traditional religion.

The traditionalists broadcasted the campaign on radio and TV stations throughout Nigeria and distributed content on social media platforms with the hashtag #EsuIsNotSatan to raise awareness about the walk:

Will you be there? #esuisnotsatan #december24#africashrine pic.twitter.com/ch8JP6gbLl

— BABA AJISEFA (@ILE_AJISEFA) December 1, 2024

Public lectures were held in Osogbo and Lagos, Nigeria. The campaign was also carried out in countries such as Mexico, Spain, the United States and other areas in the African Diaspora where there are large populations of Yoruba traditional religion adherents. 

The root of the historical error

In 1860, Bishop Samuel Ajayi Crowther, the first African bishop in the Anglican church, was commissioned by the Christian Mission Society to translate the holy bible into the Yoruba language so that the Yoruba people could access the gospel in their native tongue.

The task was crucial, as he needed to create a written form of a language that had previously existed primarily in oral form. Bishop Crowther worked throughout his life to standardize the Yoruba language and other African languages such as Nupe and Igbo. He not only succeeded in translating the bible; he also compiled the first Yoruba dictionary and published his Yoruba grammar book in 1843.

While he was translating the bible, Crowther committed an error that has caused religious discrimination for centuries. He localized Satan as Esu. This localization portrays the Yoruba deity as an embodiment of evil, like Satan. Unlike Abrahamic religions, where evil and good are often portrayed as opposing forces, the Yoruba worldview sees them as complementary. Esu is the deity in charge of controlling and regulating world forces. The Yorubas believe nothing is entirely good nor bad; and that good things might come out of bad occurrences and bad things might come out of good occurrences. 

Curbing the misconception

Even 140 years after Crowther's complete translation of the Yoruba bible was published, the misconception about Esu being Satan is still common. Popular culture and media have also promoted this negative narrative. Seeing the negative impact it is having on their religion, a group of traditionalists started a campaign in 2014 to correct the error by designating December 24 as a day for educating the public on the Yoruba religion and culture. 

Responding in a WhatsApp conversation with Global Voices on December 28, 2024, Akanbi Ifadola Afofun, a traditionalist and entrepreneur living in Lagos, Nigeria, shared insights on the misconception about Esu and #EsuIsNotSatan campaign. 

Abdulrosheed Fadipe (AF): How do you feel about the portrayal of Esu in popular culture and media?

Ifadola Afofun (IA): I feel sorry for Yoruba people because of the way Esu is portrayed in the media. Part of what I am doing is actively educating people on the identity and importance of Esu so they can understand Esu is not similar to Satan. Esu is the deity that guides the entrances, exits, crossroads, and markets in Yorubaland. It is described as neither completely good nor evil. Esu punishes those who break cosmic laws. In fact, he is regarded as the messenger of Orunmila, an Orisha that brought the word of Olodumare to the world. We are trying to stop the misrepresentation of Esu in the media.

AF: Could you share your personal experiences with the misconception of Esu as Satan?

IA: Growing up, I faced misconceptions about Esu due to the conflation of Yoruba spirituality with Western religious concepts. Esu was often wrongly equated with Satan, a misunderstanding that negatively impacted my relationship with this revered deity.

AF: What are your core beliefs about Esu's role in Yoruba cosmology? How does Esu differ from the concept of Satan in Abrahamic religions?

IA: I'm not a Bible scholar, but looking at how the bible describes Satan as the one that always fights the Abrahamic God, I believe there is nothing that Esu has in common with Satan of the Abrahamic religion. Esu is not fighting with a higher power for supremacy. Our own Esu is the figure that maintains law and order In the cosmic realm. We call Esu the police. It was the mistake of Bishop Ajayi Crowther, the man who translated the bible into Yoruba that caused the misconception. 

AF: What inspired you to participate in the #EsuIsNotSatan campaign?

IA: What inspired me to participate in the campaign is that I feel that the translation of many Yoruba concepts in English are inaccurate and misleading. Another thing that inspired me is the urgent need for the preservation of Yoruba language and culture. Another thing is I am an Ifa priest from a lineage of Esu devotees. 

AF: How has the campaign impacted the perception of Yoruba religion and culture, both within Nigeria and in the diaspora?

IA: The #EsuIsNotSatan campaign has had a significant impact on Yoruba religion and culture. A good example of the positive impact of the campaign is how Google, the most used search engine, has changed its representation and translation of Esu. I must mention the name of the man who started the campaign ten years ago at this point, Oluwo Solagbade Popoola. Since the beginning of the campaign, there has been a huge increase in the number of people who are converting to the Yoruba traditional religion. The hashtag #EsuIsNotSatan was the most trending on Twitter on December 24. Many people are now seeking to reconnect with their roots. The campaign has also opened the eyes of Yoruba traditional religion adherents to the importance of promoting their faith.

AF: What are the challenges you've faced in raising awareness about this issue?

IA: It has not been easy. One of the challenges is the lack of government support in incorporating accurate information about Yoruba belief systems into the education curriculum. This has allowed misconceptions to be introduced to the younger generation. While social media has been a useful tool, addressing the root of the problem, the curriculum remains a significant challenge. We are working towards changing the educational curriculum to accurately represent Yoruba worldviews.

AF: How do you think the campaign has influenced interfaith dialogue and religious tolerance?

IA: Some of the adherents of Abrahamic religions are becoming more tolerant towards traditionalists. Some of the preachers have stopped representing Satan as Esu. We have seen an increase in how people embrace Yoruba culture in their use of language and mode of dressing. We understand that misconception cannot be changed in a day, but we are gradually getting there.

Here is a video of the traditionalists during the walk in Lagos. 

The traditionalists are hoping this public sensitization will help reduce the misconceptions and influence religious tolerance.

Originally published on Global Voices

Screenshot of some of Nigeria's women politicians from YouTube video, ‘2023 Elections: Trials, Triumphs Of Nigerian Women In Politics | Gist Nigeria‘ by Channels Television. Fair use.

As more African nations seek to encourage gender parity in government positions, some African countries are struggling to fully integrate women into the political sphere. A recent report by the Policy and Legacy Advocacy Centre (PLAC), an independent, non-partisan, non-profit capacity-building organization that works to strengthen democratic governance and citizens’ participation in Nigeria, revealed that Nigeria ranks lowest among Sub-Saharan African countries in terms of women’s representation in parliament.

Out of the 185 countries analyzed in the report, Rwanda ranked highest in Africa, with women making up 61.3 percent of its parliament, while Nigeria ranked 180th, with only 4.4 percent female representation.

The data shows that Nigeria's 360-member House of Representatives currently has 16 women, and the 109-member Senate has just four women. Furthermore, 14 of Nigeria's 36 states have no female state legislators. Even in states with female legislators, their numbers remain disproportionately low. The situation is similarly bleak for elective executive positions: since the return to democratic rule in 1999, no woman has been elected president, vice president, or governor in Nigeria.

This low level of representation fails to reflect the population of women in Nigeria. In fact, statistics from the World Bank have shown that women and girls constitute half of the population of the most populous country in Africa.

In alignment with the principles and guidelines of the International Bill of Human Rights, the UN Convention on the Elimination of all Forms of Discrimination Against Women, the International Conference on Population and Development Programme of Action, and the 2030 Agenda for Sustainable Development, the Nigerian government has made efforts toward gender equality by formulating the National Gender Policy in 2006, which promotes a 35 percent affirmative action for women in governance processes.

Despite this policy and numerous other efforts that have been made, Nigeria has maintained the lowest rate of women's parliamentary representation in Africa, with a steady decline in numbers since 2011. The reasons behind this underrepresentation of women can be attributed to unfavorable political systems and socio-cultural practices.

Unfavorable political systems

The majority of African countries with a high or proportional representation of women in parliament have adopted constitutional, legislative, and party quotas to ensure women have a fair chance of being elected into office. However, this is not the case in some Anglophone countries in Africa, like Nigeria and Ghana, which use the plurality-majority electoral system. This system is less favorable to women candidates, as political parties win and allocate legislative seats based on the proportion of votes they receive in elections.

Male politicians dominate leadership roles within most political parties in Nigeria. These political parties usually fail to create a supportive environment or infrastructure for their women members. Hence, these women face difficulties in getting party nominations to run for office during the general election.

In Nigeria, the few women who manage to overcome the hurdles of gender discrimination usually cannot meet the financial obligations that come with vying for political offices in Nigeria since elections are highly monetized in the country. Only those who have powerful support among the party leaders and abundant financial resources to donate “gifts” to the party stakeholders are nominated as candidates.

Ayisha Osori, writer, lawyer, and social justice advocate, who ran for the People's Democratic Party's ticket to Nigeria's House of Representatives in 2015 and lost, said this about low women representation in Nigerian politics: “Elections in Nigeria are monetized and transactional, and women are already socially disadvantaged considering that in Nigeria, the fastest way to be rich is to be in government… If women are not in politics, then they cannot raise money, and if they cannot raise money, then they cannot be in politics.”

Another important issue that bars women from participating is electoral gender-based violence. Nigerian elections are often marred with violence, and women candidates and voters are always among the victims of various forms of politically-motivated gender-based violence.

One of the survivors of electoral gender-based violence during the 2023 presidential elections is Jennifer Bina Ifedi, who returned to her polling unit to cast her vote after she was attacked and injured by political thugs in Lagos. There are many more examples of gender-based electoral violence targeting women that are unreported due to inadequate media coverage.

Cultural and societal barriers

Although Nigeria's constitution grants women the right to vote and be elected, many cultural and societal barriers still prevent them from participating fully in politics and holding leadership positions. In various ethnic groups across the country, conservative beliefs persist that cast women as unequal to men and unfit for leadership roles.

Some religious practices in Nigeria also oppose gender equality, limiting women's opportunities to hold leadership positions. These socio-cultural norms have fueled the widespread belief that women, regardless of their skills or education, are primarily suited for domestic roles. For many, a woman's decision to run for public office is seen as defying societal expectations, making it a challenging and audacious pursuit.

To address this underrepresentation, experts have urged the National Assembly to enact the Gender and Equal Opportunities Bill sponsored by Senator Abiodun Olujimi of the Peoples Democratic Party to “promote equal opportunities and counter sociocultural barriers that prevent women and girls from reaching their full potential.” This bill proposes a 35 percent quota for women in both federal and state legislative seats, as well as party leadership positions. The Inter-Parliamentary Union has also called on Nigerian stakeholders to “use the current electoral process to raise public awareness about the importance of women’s participation in elections, and in parliamentary membership and political leadership.”

Originally published on Global Voices

Indigenous children in Tanzania using a laptop. Image by Kureng Workx, from Pexels (Free to use).

The tech giant Google announced that 15 more African languages would be added to Voice Search, Gboard talk-to-type, and Translate dictation on October 28. The languages included are Chichewa, Hausa, Igbo, Kikuyu, Oromo, Rundi, Shona, Somali, South Ndebele, Swati, Tigrinya, Twi, Tswana, Nigerian Pidgin, and Yoruba.  

This significant milestone will enable 300 million more Africans to interact with the web using their voices, yet it also sheds light on the massive amount of work needed to close the digital gap between Africa and the rest of the world. 

The African continent is the second most populous continent in the world, with a population of over 1.34 billion people. A study published by Statista in July 2024 showed that Africa is the most multilingual continent in the world, as there are about 2,158 living languages on the continent. While about 25 percent (520) of these languages are spoken in Nigeria, 277 and 214 languages are spoken in Cameroon and the Democratic Republic of Congo, respectively. The least linguistically diverse country in Africa is Seychelles, where only four languages are spoken.

Following the announcement, Alamazan Jak, a user from Busoga, Uganda, wrote the following comment in response to the announcement in Google's translation community:

Congratulations on the exciting announcement! Expanding dictation to 13 African languages is a significant milestone. Your dedication to inclusivity and diversity is truly commendable.

However, I couldn't help but notice that Lusoga, a vital language spoken in Eastern Uganda, is not among the listed languages. As a proud speaker of Lusoga, I eagerly await its inclusion.

Could you please share any plans or timelines for adding Lusoga to Google Translate's supported languages? Your consideration would mean a great deal to the Lusoga-speaking community.

Thank you for your tireless efforts in breaking language barriers. I look forward to your response. 

This comment shows that Lusoga and thousands of other African languages are still excluded, even as users are hungry for greater language representation. Lusoga (Soga) is a language spoken by over 3 million people in Uganda.

As artificial intelligence (AI) technologies continue to develop and gain traction around the world, they are being integrated into many tools to automate daily tasks, in many cases making people's lives and work easier. Although the adoption of AI tools is growing in Africa every day, many users cannot access them in their languages yet. If languages like Soga are not included in the development of these tools, hundreds of millions of Africans may not benefit from the limitless opportunities offered by AI. This exclusion will further exacerbate the technology gap between Africa and the rest of the world. 

The digital language barrier can impede the economic growth of many African countries by hindering speakers of Indigenous languages from accessing job opportunities and online markets. Excluding African Indigenous languages from the development of AI tools used in schools could negatively impact the education systems of many nations. Meanwhile, the use of AI in education across the continent remains as low as 12 percent.

Opportunities for language inclusion in Africa

Agriculture is the pillar of Africa’s economy. The agricultural sector is a major contributor to the GDP of many African countries. Over 43 percent of Africa’s economically active population eke out their living from farming in remote areas. However, many African smallholder farmers are facing challenges such as unpredictable weather, product insecurity, inadequate logistic infrastructure, limited access to formal financial services, limited access to digital marketplaces, economic downturns, illiteracy of many farmers, and more. The use of AI-powered digital technologies has proven effective in mitigating these challenges. However, the majority of these tools are not accessible in the farmers’ local languages.

Africa boasts many cultural heritage and diverse traditions. Language is the bedrock of these rich heritage and traditions. Languages serve as reservoirs of knowledge and traditions, but they are also the medium through which the cultural heritage of a group of people is preserved and passed from one generation to another.

Seeing the alarming rate at which languages are dying, the United Nations (UN) proclaimed a decade to protect endangered languages and preserve humanity’s common heritage in 2022. About 523 of the 3,000 dying and endangered languages that will become extinct in the world by the end of the 21st century are spoken in Africa. The exclusion of these vulnerable languages from the development of digital technologies will accelerate the threat posed by extinction. 

Many African entrepreneurs, developers, and researchers have risen to the occasion, implementing initiatives that integrate African languages into technologies. A good example is a startup in Nigeria, Awarri, working on creating Nigeria’s first multilingual Large Language Model. A Kenyan startup, Code Vast, is also using AI to provide fast healthcare guidance in local Kenyan languages like Swahili, Meru, Kikuyu, and Luo. Lelapa AI is developing machine learning tools that are specifically designed for African contexts. Another Nigerian Startup, Crop2Cash, created an automated phone system technology called FarmAdvice. The AI-powered device allows smallholder farmers to access real-time, personalized agricultural knowledge and extension services in their local languages by dialing a toll-free number on any type of device, 24/7. 

While all these efforts are ongoing to bridge this language gap, there is still a need to integrate more languages, especially those spoken in regions where there is no ongoing or little effort to bridge the digital divide. Tech giants such as Meta, Amazon, Uber, IBM, AWS and other organizations whose technologies are widely used in Africa need to get involved by also prioritizing the integration of minority African languages in their product development. There is a need for concerted efforts between local startups and government bodies to create local technologies in order to reach the areas that foreign companies do not cover.