Embedded Files
Product Background
Product Background
We Speak! is an adaptable and scalable language learning board game first created and evaluated in Tanzania. It has since been expanded in Kenya, India, and the United States and is currently available in five languages to over 7,500 students in 15 schools.
A primary (but not exclusive) focus on specific classroom contexts.
A primary (but not exclusive) focus on specific classroom contexts.
Per UNESCO, 40% of the world's population receives an education in languages other than their mother tongue. We Speak! provides a fun, engaging, and collaborative way to tackle this language barrier, or to learn a language for any reason.
Designed for seamless adaptability and scalability.
Designed for seamless adaptability and scalability.
The objective is to provide a template teachers can use to incorporate any curriculum for any language or grade level. "What if all you had to do was put a lesson in a spreadsheet, and you have a ready-to-go game to get your students engaged in learning?"
We Speak! is already available in English, Swahili, Portuguese, Hindi, and Amharic, with even more on the way! If a teacher wants to use the game but their language isn't available yet, it takes less than 30 minutes to translate everything they need.
Despite this easy adaptation and scalability, I still believe in collaboration with teachers and students to make sure the game works for them in their setting. This page discusses how I've scaled the game to different settings around the world while still working closely with partners to uniquely tailor it to meet their needs.
My Role: We Speak! is a solo project, however it was made possible through great partnerships. Please see a list of partners at the bottom of this page.
How to Play
How to Play
Play individually or as a team, and take turns rolling the dice and moving along a path. You'll encounter ten checkpoints where you must stop and complete language learning exercises such as discussing literature, telling a story, and more. There are both easy and hard exercises, with a correct "hard" answer allowing access to a shortcut. There are also "Mystery Cards" that can give your team a boost or slow down an opponent. The first one to reach "The End" is the winner!
Want to learn more? Read the full game rules, or try it out yourself on Tabletopia!
Testimonials
Testimonials
"For once, the students are actively engaged in the lessons."
– A secondary school teacher in Kenya.
"Getting students to do their reading is a problem. This (game) will encourage them to do it."
– A secondary school teacher in Kenya.
"When can we play the game again?"
– Too many playtesting participants to count!
We Speak Kiingereza: Tanzania (May 2024)
We Speak Kiingereza: Tanzania (May 2024)
We Speak! was first created in Tanzania under Michigan State's Designing Technology Solutions for Schools in Africa program.
The role of language in the educational system.
The role of language in the educational system.
Although only the native language of 10% of Tanzanians, Swahili is the primary school language of instruction, leading to 90% of the country's population speaking it (Tomedes, 2024). English is the secondary school language of instruction.
The initial reason behind the game.
The initial reason behind the game.
I worked with two secondary schools over 17 days. I wanted to create an English learning game for Form 1 (freshman) students after learning of the high dropout rates at this age due to a language of instruction transition – 1 in 17 Form 1 students dropped out of school in 2023 (Tanzanian President's Office, since unpublished).
Photo courtesy of Dr. Jenny Olson.
The analog prototype with 100% local materials.
The design process, with a retrospective.
The design process, with a retrospective.
Before we proceed, I'll be honest: I got ahead of myself. I wanted a shipped product in 17 days, and although that was achieved I realized it was still just the start. The game has since evolved and expanded, but let's start from the beginning.
Background research involved casual conversations upon arriving, then I took some initial ideas and ran with them. I decided to have exercises consist of translating full sentences from Swahili to prompt students to speak in English. Past and future tenses were a focus of the content based on teacher feedback.
The rules and structure were decided on, and an initial plan to make the game exclusively digital was shifted to an analog medium after seeing and experiencing the unreliability of the internet connections in the schools.
In hindsight, I would come to realize that much more involvement and collaboration with teachers on the educational content would be necessary, an obvious oversight (due to feeling the pressure of time) I now see with more experience. More on this later.
The original cards featured only full sentences to translate.
Playtesting: The game can work, but this is just a starting point.
Playtesting: The game can work, but this is just a starting point.
Playtesting took place at two secondary schools and involved 34 participants. 25 chose to play the game while nine chose to observe. Due to scheduling, most participants were older than the Form 1 target group.
Both sessions featured three teams with up to five players each. For the first session, I was very hands-on in guiding the players as well as determining correct and incorrect answers. I learned from this experience to be more hands-off and put moderation in the hands of the teachers.
Despite this, it wasn't too late to gain some important insight:
Students find the game fun.
The game provides a way to improve learning engagement.
I discovered a year later that this engagement was lasting – not only were students still playing the game, but they were expanding it into competitive tournaments with other students!
This means that the game has great potential, but the approach with the educational content must change. More on this later as well.
Photo courtesy of Dr. Jenny Olson.
What I'd Do Differently: Involve context experts as early as possible to ensure you start out on the right track.
Falamos Portuguese and We Speak Inglês: United States (April 2025)
Falamos Portuguese and We Speak Inglês: United States (April 2025)
Expanding to different contexts, and the start of a mindset shift.
Expanding to different contexts, and the start of a mindset shift.
Having minored in Portuguese as an undergraduate and having returned from Tanzania with the excitement of this project in mind, I decided to create an adaptation of the game.
What if the game could help bring players learning each other's languages together for a collaborative learning experience?
I reconnected with a former professor, Dr. Saulo Gouveia, to share the idea and was eventually able to playtest it with students.
This was when the switch turned on regarding several things:
There can be more to a language game than just translating!
I can better implement co-design both for language learning and to solve interaction issues like keeping players on the correct path.
With this, I was now ready to take the game into gear.
Portuguese students playtest the game with a professor.
What I'd Do Differently: Spend more time collaborating face-to-face with instructors.
We Speak Kiingereza: Kenya (August 2025)
We Speak Kiingereza: Kenya (August 2025)
I was fortunate to receive the opportunity to travel to Kenya for two weeks to continue testing and developing this game, plus to evaluate a tablet learning platform called Mind Leap, thanks to generous funding from the Mozilla Foundation.
Working with teachers to start the co-design process early this time.
Working with teachers to start the co-design process early this time.
Based on feedback from testing the Portuguese game, I presented five new types of learning exercises to teachers in Kenya, including listening and fill-in-the-blank.
I had the opportunity to work with both a primary school teacher and a secondary school teacher.
Leading up to the visit and following playtesting, teachers suggested ten more types of exercises, bringing the total to 16. Suggestions included reading a full paragraph and open-ended questions about literature.
Cards now involve reading, literature discussions, and grammar concepts such as collective nouns.
An additional change leading into the visit was to decrease the distance between checkpoints so players spend more time practicing the language than just moving along the board.
A novel idea to increase flexibility and reduce costs.
A novel idea to increase flexibility and reduce costs.
A colleague, Kennedy Nyoro, suggested removing actual questions from cards and instead only printing the question type. Teachers could use external resources for questions, allowing an essentially infinite question pool with minimal printing.
The primary school teacher opted for this while the secondary school teacher still prefers questions directly on cards. This is important: tailor the game for each classroom so that it best meets their individual needs. It also allows for quicker and easier scaling and distribution, while retaining the option for a more hands-on, personalized approach. Teachers I've worked with on later versions have also opted for different exercise types, so tailoring happens in this area as well.
What I learned about co-design from the teachers.
What I learned about co-design from the teachers.
A few things stood out to me from working with these two teachers, who had different approaches:
Translation is not part of the curriculum at all! This exercise was removed for secondary schools, while the primary school teacher felt they were still fun and chose to keep them.
It's important to build relationships with co-designers to let feedback flow comfortably. The secondary school teacher was comfortable giving feedback from the start, while the primary school teacher was initially open about some feedback but took more time to give feedback in other areas.
A group of young students playtesting the game.
Playtesting: Gaining insight from diverse student perspectives.
Playtesting: Gaining insight from diverse student perspectives.
This visit provided the opportunity to do in-depth playtesting in the African context. It was tested seven times by 41 students: 26 boys and 15 girls. The group was diverse, ranging in age from 4 to 19, with six tribes/mother tongues represented (Kikuyu and Kalenjin being the most common). Despite this, almost all students speak primarily Swahili in the home. At school, Swahili and English are the most spoken languages.
Different groups, different contexts and dynamics.
Different groups, different contexts and dynamics.
All groups were mixed in at least one way: age, gender, school, and mother tongue plus team and individual. Tests also took place both with and without a teacher, including in a home setting, to understand different ways it may fit into students' lives.
Tests followed a structure of introductions and consent, demographic and background interviews, reading of the rules in Swahili, playing the game, a focus group interview discussion, and an anonymous Likert scale survey.
Lots of great insight.
Lots of great insight.
Students were able to understand the rules and format quickly by playing the game, although there were issues with the written rules so they were therefore simplified.
Important context about class structure led to some changes:
Classes are 40 to 45 minutes depending on the school. While students generally wanted the game to last longer, time in the classroom is tight. A one minute time limit to answer questions was imposed. However, if a wrong or delayed answer is given, other teams should have the opportunity to answer and move forward if correct.
There are 50 to 60 students per class, and teachers prefer groups of ten students. This means five or six game copies are needed to accommodate each class. Additionally, to avoid overcrowding, the board size needed to be increased.
Young students can be co-designers too!
Young students can be co-designers too!
An interaction issue dating back to the original version in Tanzania has been players wandering off the set path. Each phase saw attempts to correct this, but the issue persisted. Specific questions were included in the playtesting protocol to collect design suggestions from participants to resolve this issue.
I tried things like enlarging arrows and labeling paths, and eventually joined arrows and paths into singular shapes.
Getting the game to students in schools and at home.
Getting the game to students in schools and at home.
An important part of the game is its focus on the local context. It should be able to be reproduced locally, and should involve local people.
Copies were printed in Nairobi by John Kangu (pictured), and additional ones can be created by sending editable and print-ready files to teachers to allow replication and expansion in partnership with local craftsmen.
Locally-made copies of the game are currently available in 15 schools, reaching a total student population of over 7,500. Some teachers and students were also interested in playing it online, which they can now do on Tabletopia.
A local artist, Tony Eshikumo ("Daddo"), also created illustrations to give the cards a more lively and playful appearance and to provide a visual cue as to the purpose of each learning exercise.
A copy of the board produced in Nairobi.
Additional research insights, and what to explore further.
Additional research insights, and what to explore further.
The cutoff for the version of the game tested is age 7 – younger kids did not understand the game and were not engaged. New card types were created to focus on and appeal to younger players, but I have not yet had a chance to test this. I'd like to test both exclusively with kids age 6 and under, as well as with kids in this age range playing alongside older children.
Another focus of the research was on gender dynamics. I found that neither gender was more assertive than the other within teams, instead the team leader (whether a boy or a girl) tended to be whichever player was oldest or the best at English. The big difference was how boys reacted to losing the game when playing against other boys compared to girls playing against other girls. While the boys on the losing team got quite upset, the girls moved everyone's pawns to the end and declared "everybody's a winner." This dynamic is something that warrants further research – are there ways to "lessen the blow" of losing the game for boys?
Additionally, students are also interested in playing and learning at home. While copies can be sent to homes, how will response moderation be handled without a teacher present? I'm beginning to explore things like digital flashcards for this.
What I'd Do Differently: Probe, probe, probe during interviews.
Continuing to Explore the Game's Flexibility
Continuing to Explore the Game's Flexibility
We Speak अंग्रेज़ी: Bringing the game to India.
We Speak अंग्रेज़ी: Bringing the game to India.
In partnership with Unstructured Studio, specifically the studio's co-founder Srishti Sethi, a new adaptation of the game was made: We Speak अंग्रेज़ी (Angrezee/English).
Per the Census of India 2011, Hindi is the most common mother tongue with 528 million speakers (43.6%), while English is the mother tongue of less than 1 in 10,000 people. The three language formula policy means most Indians must know at least three languages (often including Hindi and English) to succeed in schooling.
Remote playtesting of the game took place with 29 students in October 2025. While more analysis is still to come, initial feedback involved learning exercises not yet evaluated plus insight into logistical setup such as content preparation and the classroom environment while handled remotely. Initial reactions were positive, and the teacher stated she plans on making the game a weekly activity.
Unstructured Studio remotely playtested the game.
Students in Tanzania can now play the new and improved, co-designed version of the game.
Reconnecting with the original Tanzanian context – big ideas for evolution in We Speak Kiingereza!
Reconnecting with the original Tanzanian context – big ideas for evolution in We Speak Kiingereza!
In July 2025, I was able to reconnect with the main English Teacher I worked with in Tanzania a year earlier. The game had taken off, and he wanted to expand and adjust it to better meet his students' educational needs as their learning had progressed.
I was grateful for this opportunity to circle back to the original context (even remotely) to allow these students to continue enjoying the game, but with the right person (the teacher) in charge of its educational direction.
As in Kenya, this teacher also chose to focus on discussions about literature – allowing natural discussion in English beyond translation and even grammar.
Other teachers have seen the game being used in English classes and are interested in using it for their science classes. This is something that is on the horizon. Additionally, even more local schools will also be receiving copies of We Speak Kiingereza in 2026 with assistance from the MSU Tanzania Partnership Program.
Streamlining the adaption process.
Streamlining the adaption process.
To allow the game to scale, it's important that teachers can incorporate their curriculum in as low-barrier a way as possible. In March 2026, I had the chance to playtest the game in two Portuguese classes at Michigan State University.
I wanted to focus on two things. First, the number of exercises (16) seemed to be overwhelming for teachers. Instead, I focused on just four: vocabulary, correct the error, discussion, and listen and write. Next, I also tried out digital flashcards using Flippity. This way, all teachers need to do is put a lesson into a spreadsheet, then Flippity creates the flashcards for them and the game is ready to go.
The digital flashcards worked out great! It was very easy for the professors I worked with to create them with the spreadsheet, then display them during gameplay. This also would be great to let students to play the game at home without a teacher!
I'd ideally like a few more features though such as categorization and stylization, so it would be great to create my own flashcard tool. When thinking about this in Tanzania or Kenya, it would ideally work offline in cases of internet outages.
Digital flashcards can be created in seconds by just copying lessons into a spreadsheet.
What I'd Do Differently: Put a big effort into teacher guides to ensure seamless adoption (something I've been working on recently!).
What's Next?
What's Next?
Future plans and possibilities for this ever-evolving project.
Future plans and possibilities for this ever-evolving project.
I plan to continue exploring digital flashcards as I work on the game in different settings. I'm also hoping to adapt it for non-language subjects soon in Tanzania!
In addition to continued work on the game, I am currently in the process of analyzing the research data from Kenya and writing both a research paper, plus I created an academic poster for my master's defense and future conferences.
As noted in the project introduction, 40% of the world's population receives an education in a language other than their mother tongue. With the game's flexible design, I am excited to try exploring it in other contexts in addition to expanding it in its current ones. A new version for teaching Turkish is in the works, as is one for teaching English to a linguistically diverse group of learners.
The game has shown great promise thus far to unlock creative learning and reduce educational barriers. I'm open to collaboration on this, so feel free to contact me if you're interested in exploring it together!
Here is a video if you'd like to see some gameplay:
Want to read more? We Speak! has been featured in the media and presented on at conferences!
Teaching English in Tanzania Through Games by Kevin Lavery, MSU College of Communication Arts and Sciences
Game developers, professors present benefits of gaming for language learning by Reiel Ghiglia, The State News
Cross-Cultural Design to Aid English Instruction in Tanzania and Kenya: A presentation on how co-design drove the game forward in Kenya, presented at the 2025 Play Make Learn Conference.
We Speak! Addressing Classroom Language Barriers in Under-Resourced Global Regions through Game-Based Learning: A presentation at how to account for contextual factors when adapting games for different settings around the world, presented at GamiCon48V in March 2026 (view slideshow).
Partners
Partners
We Speak! was developed as a solo project, however there were many co-designers and supporters who contributed to it in big ways that I would like to acknowledge.
Educational Content and Localization
Nimpher Bureta
Ayşegül Erşil
Dr. Amlaku B. Eshetie
Dr. Saulo Gouveia
James Kanuth
Regiane Lima de Paula
Callen Mankone
Jastin Mlange
Maureen Ndelema
Dr. Deogratias Ngonyani
Josephine Nguta
Steven Odembo
Emmanuel dos Santos Machado
Srishti Sethi
Tina Wang
Camilla Zamboni
Materials Production and Artwork
Tony Eshikumo (“Daddo”)
John Kangu
Mto wa Mbu Woodworkers
Tanzania Study Abroad Support
Gryson Brown
Barry Carr
Melock Kispan
Lucinde
Omushaija Ta Mwesigwa (“Alex”)
MS-TCDC
Michigan State University Support
Rebekah Blesing
Jennifer Brewer
Susan Bonner
Jayson Burnham
Fumi Cheever
Dr. Jonathan Choti
Dr. Amanda Cote
Chris Currier
Kennedy Ellis
Grace Ekeoma Michael
Erik Goodman
Dr. Jean Hardy
Dr. Isaac Kalumbu
Mariko Kawaguchi
Tyler Lindquist
Jonah Magar
Abdul Mutashobya
Ashley O'Brien
Dr. Jennifer Olson
Dr. Fred Poole
Dr. Galen Sibanda
Claire Smith
Lee Sullivan
Owen VanderWeele
Dr. Susan Wyche
Dr. Kwame Yeboah
African Studies Center
Office for International Students and Scholars
Portuguese Club
Tanzania Partnership Program
Kenya Support
Veronicah Bahati
Moses Bundi
Dr. Miriam Charimbu
Dr. Paul Kimurto
Kennedy Nyoro
Melvin Ochami
Ogutu Omori (“Peter”)
Ramon Thumbi
India Support
Amba Foundation
Srishti Sethi
Unstructured Studio
Others
Playtesting Participants
Funding
ComArtSci Scholarship
Mark Levy ICT4D Award
Mozilla Foundation
MSU Tanzania Partnership Program
Thomas F. Baldwin Fellowship
Unstructured Studio
Timeline: May 2024 to Present
Timeline: May 2024 to Present
Page updated
Google Sites
Report abuse