An approach to the study of apps...
Duolingo is a freemium language learning platform. Its slogan is: Free language education for the world.
The main page has an icon to signify which language course(s) you are subscribed to. The page has a pyramid of icons called “Basic”, “Phrases”, “Food” etc to describe collections of vocabulary to be accumulated by the user. The bottom panel of the app allows you to navigate between “Learn” (main page), “Health”, “Bots”, “Clubs”, and “Shop”.
The app utilises machine learning with its bots, which encourage you to have a text conversation in your chosen language, and any words you don’t recognise can be clicked on and explained. It’s quite intuitive-feeling and encourages you to conjugate sentences, which is good practice. The bots have notification icons that crop up over time to remind you to chat to the bots. The app itself also sends badge notifications each day to remind you to practice.
Duolingo’s mandate is to provide free education, however the Shop function allows you to buy “health” top ups using your accumulated or purchased “gems”. For example, you can buy a Streak Freeze, which “allows your streak to remain in place for one full day of inactivity”, or a Double or Nothing which will “double your 50 gem wager by maintaining a 7 day streak”. The monetised part of the Shop function is where you can spend between $2.99 for 400 gems and $159.99 for 35000 gems. Your collection of “gems” is prominently placed in the top panel, which also has a fluency metre and a streak calendar. In the context of language learning, these tools can be really valuable as they encourage users to return regularly and practice, which is the most important behaviour to cultivate in learning a language. Where many social media apps use these methods to create obsessive use (which at its best can be a waste of time and at its worst, is a potential threat to mental health), it seems like Duolingo is using these incentives and rewards in a generally positive way, if occasionally using guilt to cultivate regular use. The “Club” function encourages users to join different online groups, which encourages community and an opportunity to practice.
Symbolic representation uncovers some interesting discoveries. There are 23 languages from which you can approach the app, and 45 languages you can choose to learn. There is an interesting bias towards Western languages, including Welsh, Norwegian and Irish, arguably not languages that should be included before languages that are much more widely used such as Hindi and Arabic, which are both languages you can learn from but not languages you can learn. I find it interesting that Hebrew is available, but not Arabic, which may suggest some geopolitical and socio-political biases. It is also possible that the developers are simply prioritising languages based on user demand.
The design of Duolingo is cartoonish and playful, which suits its gamification of language learning. I would be curious about who their target demographic was in designing the app, as it seems more appropriate for a younger audience, which is not necessarily indicative of its user base.
There is an inherent expectation that its users can see and hear, so there is an able-bodied bias (there might be a way to create dictation for blind users, for example, but I can’t find it) which is evident through its user interface arrangement. There is also an expectation of literacy, though if you are using a smart phone, that has an implicit literacy bias (as does The Walkthrough Method itself).
