Mobile Language Learning

CELSTEC works on approaches on how to support learning in context. We have developed several systems involving sensor technology, mobile end user devices, and instructional designs for blended learning scenarios. In the Learning Media Lab we have an infrastructure to prototype and showcase new solutions easily and build prototypes. Language is a typical example of something that is widely used across contexts. below we illustrate a recent example we work on.

 Language learning takes place in different settings, for example, in a structured setting in an official language course in a educational institution, or a more unstructured, and common day-to-day setting in which language is acquired in a random manner. Additionally, the type of language learnt depends on the situation; some require informal daily speech, while other settings, i.e. business negotiations, require more formal language. Furthermore, language learning is addressed towards a certain community, most often a community of native or near-native practitioners, which uses a community-specific jargon (Petersen & Divitini, 2005). Especially, in an increasingly international world, acquiring this community-specific language becomes more and more important. Particularly, non-native speakers have a demand for just-in-time, situation-specific vocabulary to communicate in a more effective and efficient way. 

This cross-context, situation-specific, community-based, and just-in-time nature makes language learning an interesting domain to explore and illustrate the possibilities and problems of a multi-platform e-learning system. In this sense, Petersen & Divitini (2005) have identified interesting community-based scenarios that include the use of mobile devices for learning (Petersen & Divitini, 2005). More specifically, they emphasise that learning in communities is important because the students need to: (1) learn in a authentic cultural context where the local language is user and (2) practice using the language with native speakers. In addition, we feel language learning would benefit from blended learning, combining de-contextualised theoretic language lessons, with contextualised authentic learning scenarios. An example of a de-contextualised language scenario is a structured online language learning course, much like the one taught at schools that train grammar, use vocabulary lists, and structured repetition. Conversely, a contextualised scenario would tailor vocabulary- and useful phrase lists to certain situations in daily-life. Paredes et al. (2005) already demonstrated the context-aware language learning tool, LOCH, which assists learners in tasks to have to be solved by interacting with native speakers in the real-world (Paredes et al., 2005). LOCH enables learners to directly get into contact with their teacher by using PDAs. The teacher can view the learner’s locations and decide to give location-specific feedback. Moreover, the learners can create contextualised information like written annotations and pictures.

In a multi-platform learning network like the one we described, several connecting language learning scenarios can be implemented. A language learning network would include a variety of different learning communities each involved in learning a different language. Each community would consist of a heterogeneous group of native, near-native, and non-native language learners that create, possibly contextualised, multilingual learning content. The creation of learning content can furthermore be combined with community-reflection where more competent learners review the work done by novices. Furthermore, learners should be helped in finding appropriate (native) peers and a community-of-practice that would help them in their learning process; in this case, it would be interesting to couple native speakers that want to learn each other’s languages. In any way, active use of a language by discussing with native peers would be an important part of language learning in learning networks.

Next to the community learning described above, language learning would also be beneficiary to self-directed learning processes, possibly mediated with mobile devices. The developed scenarios should allow for memorisation and repetition of language constructs, help to learn from errors by self-reflection on preserved learning history, and combine de-contextualised and contextualised knowledge that result in applying the knowledge learnt. Furthermore, the learning network software should help the self-directed learner in planning, structuring, self-monitoring, and evaluation of learning. Mobile devices could mediate these processes, for example by structured delivery of learning content for memorisation and repetition (Attewell & Webster, 2005).  Another example is language learning by interaction with real-world objects. The objects are enriched with language learning content, for example a text message describing the object, or an audio fragment containing a useful phrase related to that object that can be accessed using a mobile device. Thus, the interaction with the objects and learning content in an authentic situation allows learners to learn a language. Furthermore, learners can create their own language learning content connected to objects (De Jong, Specht, & Koper, 2007).

Summarising, language learning in multi-platform learning networks include the following activities:
•    Acquiring language on the move, tailored to specific situations,
•    Active use of the language, either by communication with native peers,
•    Creation of learning content, either contextualised or de-contextualised,
•    Commenting on peers,
•    Discussion with peers,
•    Memorisation and repetition,
•    Planning, structuring, and self-monitoring of learning,
•    Learning by interaction in the real-world.
In addition, countless situations could be defined that are used to contextualise the available language content, for instance standard situations as introducing yourself, ordering at a restaurant, bargaining in a shop, etc.