I like using computer games to drill my ESL (English as a Second Language) students. However, I’d like to go further. Can I use these games teach students the language?
That’s the question I’ll explore here.
The Target Audience
Since I teach intermediate level adults, I’m interested in games that are not childish, are immersive, and use language in complex ways. However, what I have to say applies to all ages and skill levels. In fact, the same game can progressively grade to accommodate increasing skill levels.
Still, for the sake of simplicity, I will assume intermediate level adults.
Why would I use computer games as a learning tool? Because they are fun, private and focused on the student.
Being fun, the students are more likely to…
- use them
- retain what they learn
Being private, they are ideal for students who are shy, anxious about participating in class or embarrassed by the possibility of making mistakes.
Being focused on the student means s/he can use the game as much as s/he likes, use the parts s/he likes, and not worry about waiting for a student to finish or holding up the class.
How is this to be accomplished? Having a game teach a language is a tall order, but it’s easier than it seems, thanks to a man named Stephen Krashen. Krashen is a linguist and educational researcher who developed a theory of language acquisition that claimed the key to learning a language was large amounts of comprehensible input rather than production of the language.
How does this help us? Well, the biggest can of worms in a language learning game is deciding how to test students on production. This would require allowing a wide array of forms, deciding how much to correct, and being able to handle semantics. However, providing large amounts of comprehensible input is a lot easier.
What is comprehensible input? It’s audio and text that’s slightly above the skill level of the student. This means s/he will need some contextual aids to understand the language, and this comes from the animations, graphics and additional audio in the game. With enough scenarios, the student can get plenty of input.
So what will our game be about?
Once again, Krashen comes to our aid. He (along with Tracy Terrell) states that conscious grammar study should not be taught, but rather, students should be taught topics of interest using the language they wish to pick up.
So the game should not be about language, but should be about other stuff, but it happens to use the language students need to learn. This isn’t hard, since language is how we communicate. So we’d set up puzzles, information gathering and various other tasks that the student would need to solve, and the student would get all the information needed in the language s/he is meant to acquire.
This is Task Based Language Learning, and it’s already used in classrooms across the globe. In fact, most of my classes are Task Based.
But what should these tasks be?
A task has to be interesting to the student, so one obvious source of tasks are student goals. Students want to improve their English for many reasons, including finding a job, making friends and being able to handle doctor visits.
These goals can be broken down into specific tasks. For instance, finding a job can be broken down into finding job search resources, finding jobs they qualify for, applying for the job, doing well on the interview, and so on.
Interestingly, these tasks – while related – are relatively independent and so each could become a mini-game. At best, the entire game can be a collection of independent mini-games. At worse, these mini-games are only connected by links and some shared information. Either way, it allows for modular development and a phased release of new mini-games as needed.
How Does it Look?
So let’s assume our mini-game is finding a job for which the player qualifies. The player is assigned a character, and gets a voice-over describing his/her qualifications, experience, education and strengths. Now the student is told to look for a job, and this entails several skills and tasks within the mini-game:
- Follow directions to the employment office
- Choose the right job from the counselor’s description of various jobs
- Choose a job and follow directions (GPS upgrade?) to the company
- Tell the hiring manager why s/he is there
- Enter an interview process if s/he qualifies
- Inquire as to why s/he is being turned away if s/he doesn’t qualify
Feedback is available at each step. For instance, making a wrong turn will let the student know a mistake was made. Likewise, inquiring as to why s/he is being turned away can lead to a dialog with details which will help the student know the mistake s/he made in choosing that job. This is essential for correcting mistakes or knowing if the language was correctly understood.
Is it Fun?
Finding a job. Doesn’t sound riveting, does it? Well, many of the Sims had people doing mundane things, and look at how successful they were (more on the Sims later). Also, it isn’t just the material, but how it’s handled. For instance, the game could be more fun by adding some of the following game elements to it:
- Competition (vs. a timer, computer player or previous high score)
- Humorous feedback
- A progress meter showing how close to the goal the student is
- Emphasize intrinsically interesting aspects of the task, like:
- Resource Management
For instance, in the job hunt mini-game, if the student takes a wrong turn while following directions, s/he can meet with an unpleasant, yet humorous fate, like being mauled by a pack of chihuahuas (Humorous feedback). Likewise, if the game is open-ended enough, it can provide a sort of sandbox. For instance, the student can keep going to the employment office, drive to whatever location s/he sees fit to drive to, go back to the employment office to refresh his/her memory about the other jobs, and so on (Exploration).
Score keeping can be added to penalize the student for mistakes without necessarily ending the game. For instance, if the student has limited savings (Resource Management – the job is to replenish resources), then the money can drop with every trip (representing gas, depreciation on the car, etc…).
So, that’s one way. But maybe we have a shortcut. You remember I mentioned the Sims? Well, people have used them as language teaching tools, so why can’t we do the same? Other options include MMPORGs, MMOs and Interactive Fiction.
If we could get access to the source of such a system, or a system with a flexible API, we could alter it so it becomes a language teaching vehicle. In this case, we’d want it to provide a lot more audio, more involved language forms, well formed inputs, and so on.
And speaking of input…
We want the player to focus on comprehension, but this doesn’t mean we completely ignore production. For instance, simply selecting from a menu of responses is too simplistic (except for beginners). We’d like a system in which the user must enter correct responses while exercising some grammatical competence.
We could have options in the form of scrambled sentences, with the player unscrambling the sentence that represents the option desired. Perhaps there could even be a paraphrase of the option so the player doesn’t have to deal with the dual challenge of deciding what to do, and figuring out what the scrambled sentences mean.
So for instance, back to the job interview; let’s say the player stands before the hiring manager, who asks, “How may I help you?”.
The player can then have these jumbled options:
(Ask about the job)
(Ask for a job)
The player would have to unscramble the options and then present the most suitable of the unscrambled versions. Note how this tests cultural conventions like politeness too. If the player did it wrong, the hiring manager could express confusion. If the player did it right, but asked the wrong thing, the hiring manager might kick him/her out in a humorous way (two burly guards hurling the character on the sidewalk).
The sentences could even be lexically set up, so the user puts together phrases and clauses, thus raising consciousness of phrases, idioms and other collocations:
We could go further. Once the sentence is properly formed, the player may have to pronounce it within a certain threshold of accuracy. Mispronunciation can cause confusion (although it should be clear if the player unscrambled wrongly or pronounced incorrectly – feedback should be precise enough to allow correction).
Even armed with the info we want, getting content can still be challenging. Sure, we know what happens in a job interview, but when we write it down, is our dialog stilted? Is it overly idealized? Is it overly simplistic? We may want a 3rd party source of utterances to correct our biases.
We could try searching Language Corpora for authentic content, although we need to be careful as these Corpora are often written or speeches, which may be structurally different from dialogues. We could also look for transcripts, job-training videos and other sources of authentic language materials.
Although I’m considering this system for intermediate level students, it is helpful to think of how it would scale for different levels. We can scale by adjusting a few dimensions. Namely, the more advanced the student…
- the more complex the language.
- the less time s/he gets to complete a task.
- the more accurate must his/her pronunciation be.
In fact, it may be simpler to design this for beginners first, and then work up to higher levels. Beginners need simpler language, can get by with a menu of options, and face simpler situations. They may need some more cues or translations at key points however. Once done, then greater complexity can be added for more advanced students.