Wisp and the Next Generation of Glossing Tools for Video Games
By Frederick J. Poole, Michigan State University
DOI: https://www.doi.org/10.69732/SECX9198
Introduction
Glossing tools allow learners to access quick information about a word, phrase, grammatical structure, or sentence in context. As early as the Medieval periods monks were known to add glosses (e.g. notes) to texts to help facilitate understanding of Latin scripts. Similar approaches were used in the early 19th century as language learning texts often included glosses in the form of annotations in the margin or between the lines. The goal of using these glosses is to provide linguistic support while not disrupting the reading process.
With the rise of the computer assisted language learning (CALL) field in the late 20th century, many glosses started to take a digital form. Tools like Hot Potatoes, Lextutor, Wenlin, and DimSum, allowed teachers to create digital texts that learners could interact with by clicking on or hovering over a word to get an L1 translation. Today, teachers have access to tools like ReadLang (Carson, 2022; Quinn, 2023) which allow users to provide a glossing tool to almost any webpage on the internet. CLILStore further expands these capabilities by allowing users to create glossed YouTube videos.
Even major streaming services now have extensions that allow users to convert videos into glossed experiences. Language Reactor essentially allows users to turn a TV show or movie into a scaffolded language learning experience. Such tools greatly expand what resources can be used for language learning.
Glosses are also heavily supported by research. Scholars have long argued that glossing can keep the focus on meaning over form (Poole & Sung, 2016) and can help improve reading fluency (Shen & Tsai, 2010), reading comprehension (Hong, 1997), while also lowering reading anxiety (Zhao et al., 2013).
I have often been interested in exploring how glosses could be integrated into video games. However, video games present a number of challenges compared to glossing webpages or videos. Webpages and videos have very similar structures which makes it easier to design a glossing tool for. In contrast, video games can vary substantially. They may be 2D, 2.5D (isometric games), 3D, full on virtual reality, text based, among many others. Further, in-game language might appear in cut scenes, in an in-game book/text, in an inventory, or within a dialogue. To put it simply, video games are extremely complex environments which makes creating a tool that is one-size-fits-all quite difficult.
As a result, most integrations of glossing tools into games comes in the form of educational games or software with built in glosses like Influent Language Learning Game, which provides linguistic information for objects in the virtual world when a player interacts with the object. Other software like ImmerseMe or Mondly VR attempt to place learners in real world settings (e.g. on a train, at a restaurant) and then provide support through in-game dictionaries and/or suggested in dialogue responses. The problem with many of these attempts to provide linguistic support is they often end up being decontextualized and/or become a rather fancy flashcard system, as is the case with Influent Language Learning Game. Another issue may be that the learning mechanics become so cumbersome that they interrupt the natural flow of using the language and thus also disrupt the enjoyment of using a game or software to learn, as is the case with ImmerseMe and MondlyVr (in my humble opinion).
I have also dabbled with creating glossing systems in a few games that I designed for Chinese classrooms (Poole et al, 2018; Poole et al., 2022). My first attempt (Poole et al, 2018) involved an interactive game I made using Twine (twinery.org). In the game learners could explore different environments (e.g. mountains, forest, and desert) and meet people by clicking on them. Glossing was added via dictionary, which you can see on the right side of Picture 3. When learners clicked on a word it would be added to their dictionary with a definition.
My second attempt (Poole et al., 2022) involved designing a language learning game using RPG Maker MV. The glossing system in this game only provided the transliteration, or the Roman script for the targeted Chinese characters that were identified as potentially difficult for the learners in my study. This was because these learners had strong oral skills, but struggled to recognize characters. Thus by providing phonetic information it could support their reading skills.
While I believe that these two games and uses of glossing systems were valuable and led to learning, there are two problems with my glossing tool experiments. First, both of the games that I created were designed for a specific classroom/setting (e.g. the Utah Chinese Dual Language Immersion program), and they took me a long time to develop. In other words, the first problem is that it is not efficient for every teacher or even every school district to create their own game/glossing system. The second problem is that although my learners reported enjoying the game and I liked the games, they are admittedly much less fun than many of the commercially available games out there today. It is also important to note that in today’s world we have access to thousands of commercially developed games, and those games include multiple language options.
Wisp
Wisp is one of the first tools that I have seen that has attempted to create a one-size-fits-all glossing tool for video games. Instead of creating an educational game with a glossing mechanic, Wisp allows users to add a glossing mechanic to any game that they are playing. Given that most games on Steam (a video game distribution service) have multiple language options (as noted above), this greatly increases the resources and/or opportunities for using digital games to learn one’s target language. Wisp currently boasts that it can be used with any language. However, it should be noted that the accuracy of glosses is likely tied directly to the performance of today’s large language models (like ChatGPT) in your target language. In other words, languages with high representation in large language model’s training data such as Spanish, Chinese, and/or Arabic, should provide reliable glosses. All of this being said, this tool is currently still in development and thus further testing is warranted. Finally, it is important to note that Wisp does collect learning data, but does not share this data publicly at the moment.
Name of Tool | Wisp |
URL | https://newcomergames.com/Wisp/ |
Primary Purpose of the Tool | Provide glossing for language in video games |
Cost | Currently it is completely free, but will likely include a subscription at a later time. |
Ease of Use | Simply start the program and toggle on the options that you want to apply to your video game. The video game must be in windowed mode. |
Currently, Wisp has a number of functions, called Wisps, that provide language learning support while playing a video game. The first one is called Shine. This is the most basic form of support and allows a user to get a direct translation of a word, an explanation of its use in context, as well as a pronunciation of the word.
The next Wisp (or function) is called Stream, which allows a user to drag a shaded screen over an area of text and get a longer translation for a sentence. This may be useful to understand how individual words work within larger contexts.
The next Wisp is called Refresh. This Wisp should look somewhat familiar for those who have used the Duolingo App. This function allows you to study words that you have seen within the game by building sentences and translating sentences. The key difference here is that words are contextualized within the games that a learner was playing. Thus, there are no more silly, made up sentences. Contextualizing one’s learning not only helps learners make inferences about meaning based on the context, but it also provides an example of how words are used in the real-world and may lead to better vocabulary retention (Godwin-Jones, 2018). Furthermore, learners/teachers may be able to specialize their language learning by focusing on specific genres related to certain games (e.g. farming, sports games, and so on).
The final Wisp, or function, is Chronicle. Chronicle saves words that were looked up and streamed while playing the game and allows users to review vocabulary and/or phrases after they have stopped playing.
This tool is currently free and will likely remain free for limited use in the future. While it does currently function, it is clearly still being developed and as such some glitches should be expected. However, if you are a gamer or interested in using games in your classes, I highly recommend getting familiar with the tool and following along as it continues to add new functions and becomes finalized.
Practical Applications of Glossing Tools
The primary benefit of glossing tools is that they expand what resources can be used in the L2 classroom by providing language support. Thus, text, videos, and now games that may have been too difficult, are now options as a source for language learning. That being said, integrating glossing tools into the classroom still requires important design decisions by educators.
For example, we must first consider when to use glossing tools. Adding a glossing tool to a text typically allows learners to read faster and gain comprehension more quickly, but may result in less deep processing of vocabulary and syntax. Thus, glossing tools are best used (usually) when developing reading skills are not the primary objective. They are most effective when the focus is on comprehension and/or fluency.
It is also necessary to consider personalization and linguistic support provided. If the glossing tools allow for these customizations, teachers may consider if they want to provide L1 translations or L2 explanations of targeted concepts. L1 translations are generally better for novice learners as they enhance comprehension and fluency, but L2 explanations can provide additional input and may be more appropriate for advanced learners.
The activities that come before and after the glossed text are also important. For example, if the goal is developing reading skills, we might use a glossing tool first to provide learners with a schema of the text, thus reducing cognitive load before reading it a second time. On the other hand, if a follow up activity involves a debate, discussion, or information exchange related to the text content, then a glossing tool that promotes a quick and accurate reading of the text may be sufficient. The key here is that we should consider how glossing a text fits into our lesson plan.
Finally, training learners to use glossing tools can impact the outcome significantly. In my own classrooms, I noticed that less proficient language learners tended to gloss everything (even words they knew!), whereas more proficient learners were careful to only gloss words they did not know or had doubts about. Having discussions about the effective use of glossing tools can be very helpful. Additionally, learners should be prompted to revisit words they glossed while reading.
Given that Wisp facilitates reading comprehension within games, it is important to design activities that allow us to leverage the information that learners pull out of games in the classroom. Below are some ideas on how games like Stardew Valley, Pokémon, and The Banner Saga could be leveraged for in-class activities with the help of a glossing tool like Wisp.
To provide some context for the games:
- Stardew Valley is a farming simulation game where players interact with villagers, grow crops, raise animals, and complete various tasks to build a farm.
- Pokémon is an adventure and strategy game where players capture and train creatures called Pokemon, which are then used to battle other trainers in a large open world.
- The Banner Saga is a tactical role-playing game set in a Viking-inspired world where players lead a caravan through a narrative-driven adventure filled with turn-based battles and difficult moral decisions to make.
-
Glossed Game Journals (Stardew Valley)
After playing Stardew Valley with glossed in-game texts (e.g., dialogue from NPCs, quest descriptions, or item details), learners can identify words glossed and use them to write a summary of in-game activities.
-
Collaborative Game World Mapping (Stardew Valley)
In Stardew Valley, learners use glossing to quickly gather information about different locations (e.g., farms, caves, shops) and the activities or items associated with them. In groups, learners share their findings and collaboratively create a visual map of the game world, annotating it with key details they gathered from glossed text (e.g., what crops to plant, best fishing spots).
-
NPC Report (Stardew Valley)
In Stardew Valley, learners use glossing to gather information about specific NPCs (e.g., their likes, dislikes, or role in the village). In class, they create and share NPC reports with detailed descriptions based on their glossed readings.
-
Character Analysis with Glosses (Pokémon)
Using a glossing tool in a Pokémon game, learners can gather information about specific Pokémon trainers or Pokémon abilities. In class, they could create a character profile or strategy guide, sharing key information about a particular trainer or Pokémon’s abilities. This could lead to conversations around the best Pokémon, and when and how certain Pokémon should be used.
-
Decision Debate (The Banner Saga)
In The Banner Saga, decisions heavily influence the story. Using glosses to better understand the implications of dialogue choices or decisions they make in the game, learners can engage in in-class debates, explaining their in-game decisions and justifying them using the glossed information they gathered.
Conclusion
The activities that I have put forth here were created with the classroom in mind. But I think it’s important to acknowledge that not all teachers will be interested in bringing games into the classroom given the cost and the required equipment to do so. Thus, these activities could also be used for extracurricular activities/events. Additionally, I have used gaming as a journaling activity in some of my classrooms. Students start the year by identifying a game that they want to play in the target language. They share this with the class in the form of a short video and then once a week, they write a short journal/blog entry to discuss their learning journey in the game. With novice learners these reflections often start in the L1 and then slowly shift into the L2 as learners gain proficiency. Using a tool like Wisp would greatly support an activity like this and learners could be encouraged to share words they glossed with example usage each week.
For more examples of activities that leverage games at Games2Teach and in the Ludic Language Pedagogy Journal. Using games in the classroom can help make language practice and use meaningful by providing a context and purpose, however without tools like Wisp the linguistic challenges have been a barrier. If teachers explore the use of Wisp with digital games in their classroom I would love to hear more about it!
References
Carson, A. (2022, June). ReadLang.com: Supporting semi-autonomous reading has never been so easy. The FLTMAG. https://www.doi.org/10.69732/HPWN7677
Godwin-Jones, R. (2018). Contextualized vocabulary learning. Language Learning & Technology, 22(3), 1–19. https://doi.org/10125/44651
Hong, W. (1997). Multimedia computer-assisted reading in business Chinese. Foreign Language Annals, 30(3), 335-344.
Quinn, S. (2023, January). Using ReadLang for listening. The FLTMAG. https://www.doi.org/10.69732/LJAO6340
Poole, F., Clarke-Midura, J., & Ji, S. (2022). Exploring the affordances and effectiveness of a digital game in the Chinese dual language immersion classroom. Journal of Technology and Chinese Language Teaching, 13(1), 46-73.
Poole, F., & Sung, K. (2016). A preliminary study on the effects of an E-gloss tool on incidental vocabulary learning when reading Chinese as a foreign language. Journal of Chinese Language Teachers Association. 51(3), 266–285.
Poole, F., Franco, J., & Clarke-Midura, J. (2018). Developing a personalized, educational gaming experience for young Chinese DLI learners: A design-based approach. In R. Zheng (Ed.), Digital Technologies and Instructional Design for Personalized Learning (pp. 253-274). IGI Global.
Shen, H., & Tsai, C. H. (2010). A web-based extensive reading program and its assessment system. Journal of the Chinese Language Teachers Association 45(2), 19-47.
Zhao, A., Guo, Y., & Dynia, J. (2013). Foreign language reading anxiety: Chinese as a foreign language in the United States. The Modern Language Journal, 97(3), 764-778.
Frederick, Thanks for the very helpful historical background of glossing, the detailed analysis of Wisp, and your personal reflections on why and how it all matters for learning. I have new ideas about glossing, and the courage to try it in some of the teacher training I’m doing. This is a great example of why FLTMag is such a useful and valuable publication.
Thanks Jeff for the kind words. I would love to hear more about what you do with glossing tools for teacher trainin and swap some notes at some time.