While we are charting exciting new research territory with our goal of enabling “zero to reading” with minimal skilled supervision, our efforts are firmly grounded in decades of research in language learning and sociocultural psychology. Our team includes well-established ISLA and dual language learner education researchers from UC Santa Barbara and Carnegie Mellon University — our content design draws heavily on their experience and insights.
Second language reading development is driven in large part by vocabulary knowledge (meaning of words), phonological awareness (recognition of sounds), and orthographic awareness (recognition of spelling) (Jeon & Yamashita, 2014). Understanding the meaning of written words depends to a large degree on the reader’s ability to decode spelling and corresponding sounds. Likewise, being aware of spellings and sounds contributes to the learning of new words while reading (e.g., sounding out unfamiliar words and inferring their meaning from context).
Our digital content focuses on these component skills through gamified activities and stories. We draw on the principles of sociocultural psychology to devise audiovisual scaffolding for language acquisition. The central tenet of social psychology is that learning is mediated by assistance “from the outside” (Vygotsky, 1978) — parents, caregivers, teachers, and in our case, our GLEN Learn app, support learning by directing and guiding children’s behavior. In the context of language learning, this often means drawing learners’ attention to important features of new words and grammatical structures as they engage with them by reading, writing, listening, or speaking. Importantly, however, support must be contingent on learner need — it needs to be provided when needed, but it should be withdrawn as the learner becomes capable of performing independently. Building on pioneering work in computerized dynamic assessment (Lantolf & Poehner, 2013; Leontyev, 2016; Qin & van Compernolle, in press), our content includes built-in audiovisual support that learners can access as they play the games to draw their attention to such language features as phonemes (sounds), graphemes (letters), and syntax (word order) that underpin reading development.
The audiovisual scaffolding provided by GLEN Learn links a phonological loop (remembering what is heard) with a visuo-spatial sketchpad (remembering what is seen). This helps the child take advantage of working memory capacities (Baddely, 2003). Storing word-sound-spelling matches in working memory during the games, through repetition and over time, enables storage of word-sound-spelling matches in long-term memory (Paradis, 2009). Matching images to sounds and written words leads to vocabulary acquisition, and matching sounds to letters and written words leads to reading ability.
Figure 1 depicts the learning architecture that we have developed based on the preceding insights. The curriculum around which our digital content is designed aims to develop reading abilities, starting with basic component skills — vocabulary, phonological awareness, and orthographic awareness — and leading to using reading to learn and further develop component skills.
Figure 2 shows how the GLEN Learn app maps this architecture to concrete content components (not all of which have been deployed yet). See this blog post for more details.
Since we do not rely on skilled supervision, we also aim to emulate the kinds of assistance a teacher might be able to offer students who encounter difficulty while performing a learning task, for instance, by building an interactive dictionary that runs in the background. This means that the user can “ask for help” by clicking on a relevant word or phrase, much like a student in a classroom might raise his or her hand to solicit a teacher’s assistance during individual or small-group work. Thus, explicit guidance is available when needed, but it is not automatically integrated into the gamified content. This minimizes the risk of interrupting a learner who is making progress independently and/or does not require assistance, since offering assistance when it is not needed can in fact hinder learning (Aljaafreh & Lantolf, 1994; van Compernolle, 2015; van Lier, 1988).
Finally, we recognize that we are at the cutting edge of language learning for young English learners. Thus, in addition to leveraging ISLA research, we must constantly experiment and iterate. A critical component of this approach is our user analytics (data on learner engagement and progress metrics), coupled with the qualitative feedback from teachers and deployment partners. This feedback loop will drive further research and development for continuous improvement of our content, informed by our learners and leveraging the expertise and creativity of the researchers in our team.
References:
Leontyev, V. V. (2016). Discourse 2.0: Language and New Media. Volgograd State Univ Pr-kt Universitetskii 100, Volgograd, 400062, Russia
Van Compernolle, R. A. (2015). Interaction and second language development: A Vygotskian perspective. Amsterdam: John Benjamins.van Lier, L. (1988). The classroom and the language learner. London: Longman.