Using Handhelds:

Learning a Language “Anytime, Anywhere”

Phyllis Larson

Craig Rice

St. Olaf College

Abstract

Handheld devices, also known as PDAs, are generally viewed as a technology that makes the mechanics of managing a professional life easier: providing electronic calendars, addressbooks, todo lists, and so on. What has largely been ignored is their usefulness in teaching and learning. This paper describes how a partnership between a Japanese language faculty member and an information technologist, combined with support from the Freeman Foundation, has resulted in the thoughtful, planned, and appropriate use of handheld technologies in the teaching of Japanese language. The authors document their goals, their strategies and software to address those goals, their rationale for using handhelds in pursuit of their strategies, their implementation, training, and evaluation of the effectiveness of the handhelds, and their plans for future use.

Our Goals

This handheld project developed out of a discussion between the two of us about one of the major challenges of teaching Japanese in the United States: the difficulty of learning to read and write, and the amount of class time that needs to be delegated for it. In order to read and write Japanese, a student needs to master two syllabaries of 47 items each plus variations, and nearly 200 Chinese characters (kanji), most of which have at least two readings. Adding to the difficulty is the fact that there are no cognates in English for either the characters or vocabulary students must learn in Japanese. Further, students need to learn and practice the correct order for writing the characters, so learning to write involves meticulous and patient study of the stroke order and visual aesthetics of the characters. Complicating all of this is American students’ lack of skill in memorizing for long-term retention. As we talked, other concerns surfaced: students need to be more active in their learning; and students’ individual learning styles should be addressed.

We realized that handhelds seemed promising as a way of meeting these challenges. We decided to investigate whether there was any software we might be able to use, especially any that would allow us to create lessons tailored to our specific purposes.

Our goals became these:

1) to support long-term retention of characters and vocabulary in Japanese;

2) to encourage more active student learning;

3) to address individual learning preferences; and

4) to tailor software to specific course objectives.

Our Strategies

We knew we could increase long-term retention of characters and vocabulary if we were successful in getting students to increase the frequency with which they manipulated the linguistic items and the sheer amount of time devoted to practice. Beyond that, students needed to receive immediate and accurate feedback on their responses. We hoped to find software that would allow us to implement these strategies with our own tailored materials.

Why Handhelds?

While flashcards share some of the advantages of a handheld, in that they are portable and offer randomized practice, we found that we could improve accuracy and increase interactivity with the use of the handheld. In previous classes, students had made their own character flashcards by hand, not all of them accurately. When they drilled themselves, they had no notation of correct stroke order, so the materials from which they studied did not give them enough information, and sometimes they practiced forms incorrectly without realizing it. The handheld offers students practice in a variety of formats and with immediate and accurate feedback.

By combining other things a student may want to do (keep a schedule, take notes, and even play games) the more likely students are to have the handheld with them. Because of this convenience, and the fact that the digital environment is an appealing way for students to work, the handheld lends itself to brief, more frequent practice.

For languages that utilize non-roman scripts such as Japanese and Chinese, learning how to correctly draw, both in design and in stroke order, is as important as being able to recognize and pronounce the character. Traditional computers do not support this function, since drawing a character with a mouse is not particularly effective, and entering the characters on a keyboard is even less so. In contrast, most handheld devices support some kind of “drawing” function for entering information; this combined with appropriate software can turn the handheld into an ideal practice device. For PalmOS devices, one-stroke shorthand notation called “Graffiti” is used for entering.

Unfortunately, for language learners, especially those who are learning non-roman scripts, entering letters or characters using this Graffiti system, even as it might be adapted to non-Roman character sets, is counter-productive: there is no reinforcement of the correct character drawing or stroke order; and worse, multi-stroke kanji cannot be drawn by the single, continuous stroke required by Graffiti.

We wanted to capitalize on software that allowed students to draw characters completely and correctly. Since the ability to do this is limited to specific software, this meant that students could not effectively create PalmOS standard To-Do lists and Memo items; such activities would have been ideal, providing an opportunity for real use of Japanese.

The Visors

During the fall of 2000, a variety of handheld devices were on the market. Our first decision was whether to pursue a device that supported the Palm Operating System (PalmOS) or the Windows CE operating system. Two primary factors played a role in our decision: available software and cost.

Of all the potential software we identified, more was available for the PalmOS environment than for the WindowsCE environment. Further, PalmOS devices tended to cost much less than their WindowsCE counterparts. Lastly, we found the PalmOS devices more intuitive to use.

Among the PalmOS handhelds, Handspring offered a unit called the “Visor Deluxe.” Costing only $150, this device was affordable and came standard with 8MB of memory. At the time, this much memory was quite liberal, and knowing that our Japanese software – dictionaries, fonts, and applications – might well consume more than the 4MB standard at the time, our selection of the Handspring Visor Deluxe was straightforward.

With financial support from St. Olaf College, and a generous one-for-one donation of refurbished Visors from Handspring, we equipped our fourth-semester Japanese language students with Visors in the spring of 2001.

Software

In preparation for our project, we researched software available to assist in Japanese language learning. One package, KingKanji (www.gakusoft.com; $35 per copy, site licenses available), fit our needs very well; it was flexible learning tool, allowing customizable lessons, stroke order display, and user interaction. We had initially looked at a program called Kanji Hanabi (www.neth.com) and were impressed by its intuitive user interface and its large, readable characters. Kanji Hanabi did a nice job of demonstrating stroke order but lacked the ability to display multi-kanji vocabulary or phrases; further, it was neither customizable nor did it allow students to practice drawing the characters.

KingKanji included the ability to customize lessons and provided a stroke order demonstration (the longer you held the stylus on any character, the slower it drew the character, stroke by stroke).

Students can work with any one lesson, or any group of lessons, making review of arbitrary sets of material very easy. KingKanji selects a vocabulary word or kanji, and based on student preferences, displays (or withholds display of) the definition, the reading, and the word, phrase, or kanji. The student then draws the characters in the large rectangle; at any time, the user can click on the character to review the stroke order. When finished, the student clicks on “Show” to display any of the top three boxes that are not already displayed. The software does not actually evaluate the student’s responses, so it is up to the student to determine whether the answer is right or wrong and whether to repeat the exercise.

Although we could create custom lessons, doing so with the initial versions of KingKanji was very difficult; current versions of KingKanji include a more user-friendly lesson creator that much simplifies the process. The vendor’s web site also contains downloadable lessons others have contributed, including our own.

We also worked with AvantGo (www.avantgo.com; free) in conjunction with CJK-OS (www.dyts.com; $28 per copy, site licenses available). CJK-OS is a PalmOS add-on that allows the native display of Japanese (as well as Chinese and Korean) that we had hoped would allow students to take notes and to create to-do lists in Japanese. Eventually we discovered that its ability to enter Japanese very weak if not impossible. When used in conjuction with a handheld web browser like AvantGo, we were able to provide authentic language web content, for example, articles from the Japanese newspaper Asahi Shinbun, on the handhelds. We created an AvantGo “channel” and created links within it to other web content.

We found several other software packages that could be useful. At the beginning level of Japanese language learning, the free program Kana – Hiragana Flashcard Trainer (www.delph.com/kana) does one thing very well: flashcard practice of hiragana characters. We are excited about the many functions provided by the memory-intensive Dokusha program that provides a Japanese-English dictionary, a Kanji dictionary, and several other features (www.geocities.com/andrew_brault/dokusha/index.html), and looked briefly at BDICT, JAQUIZ and JDICT.

Preparation

During the fall semester of 2000, we identified the key software packages we planned to use and prepared 24 lessons tied to the chapters from the textbook for the Intermediate Japanese course for the KingKanji program. We created several lessons containing vocabulary, phrases, and kanji for each chapter in our textbook, Genki II: An Integrated Course in Elementary Japanese.

We divided vocabulary and kanji from each lesson into between four and eight sub-lessons. We started each sub-lesson as a Microsoft Word file containing one line for each word or phrase in the form:

kanji (or vocabulary word) [pronunciation] /meaning/

for example:

国 [くに] /country/

We saved each lesson file in a special format with a special filename (“EUC Encoded Text” with a “.kf” extension), then used a utility supplied with KingKanji to create the “kusr.pdb” file which is finally loaded onto the Visor for KingKanji’s use. We found this multi-step process time-consuming and prone to error.

In the spring semester of 2001, we launched our project, distributing to each of twenty students in our Intermediate level Japanese language course a Visor with a docking cradle (used for synchronizing the handheld with a desktop computer), and a CD-ROM containing the synchronization software.

Since each license for our 20 copies of KingKanji and CJK-OS required a unique, several character user registration, we gave each Visor a distinct name: “visor001,” “visor002,” “visor003,” etc. Preparation of the Visors for distribution to students was frustratingly slow and involved the following steps:

1. Resetting each Visor to its “initial” and clean state

2. Applying a plastic screen protector

3. Setting PalmOS preferences to maximize performance

4. Installing our Japanese language learning software

5. Placing each piece of software into a “Japanese” PalmOS menu category

6. Installing individual licenses for CJK-OS and KingKanji

7. Synchronizing an initial set of AvantGo web pages

The following fall semester, we did not have to create new KingKanji lessons, and instead of treating each Visor as a unique unit, we created one “master” Visor, saved an image of this Visor to a Handspring 8MB Backup Module, then restored this image to each student Visor. This dramatically reduced the time for preparing Visors.

Student Response and Evaluation

Feedback on the Visors after the first semester was disappointingly lukewarm. Students did not find the units or the software particularly helpful in their learning of Japanese. Based on discussion with students, we learned that our brief introduction at the beginning of the semester was insufficient in helping them become comfortable with this new learning environment. This was a wake-up call for us, since we both felt students were coming to college with a strong background in things technological.

We also found that the synchronization process (“HotSync”ing data between the Visor and a desktop computer) was difficult to configure and painfully slow, AvantGo in particular. In the end, we abandoned HotSync’ing, meaning that our AvantGo-accessible web pages were not used.

We re-evaluated our training, and in future semesters spent an entire class period talking through (and working with) the Visors and the course software. In addition, we devoted fifteen minutes every other Friday to discussing how things were going. One of the most helpful aspects of this was the student-to-student sharing of how they are doing what – having them demonstrate something they have discovered, or showing others how they are doing something on the Visor.

Evaluation of successive semesters resulted in reports of extensive student use and strongly positive feedback. In a survey of 17 class members, all but one student stated that they used it any where from several times a week to several times a day. They found it saved the time they used to spend making flashcards and allowed them randomized reviews of vocabulary and kanji whenever and wherever they had a few minutes.

Limitations of the Visors

Students’ comments were extremely helpful, and several of them focused on some limitations of the KingKanji software. Students complained that, while it was very useful to see the stroke order of a character, it was often too small to read easily.

We also ran into memory limitations, forcing us to limit the software we could install on the Visors.

As we became more familiar with the capabilities of the Handspring Visor, we soon began thinking about the “what if’s.” With our goals of encouraging more active student learning and addressing individual learning preferences in mind, we started looking for ways of incorporating audio content: using the handheld to present pronunciations of vocabulary and even short dialogs. We experimented with the TotalRecall Voice Recorder (a module that plugs into the Visor and acts as an audio recorder) and with the MiniJam MP3 SpringBoard module (that allowed us to play MP3 files from the Visor). We found immediate application for these (more on this below), but they were quite expensive and did not always work reliably, especially when we inserted the module into or removed the module from the Visor.

The Next Stage

During the Fall of 2001, we evaluated the Visor project, brainstormed some additional pedagogical applications, and researched new software and handheld devices. After coming up with our ideal plan for the next stage, we submitted an application to the Freeman Foundation and received a grant to expand the use of handhelds to all of our Japanese language students and to customize software and a more advanced handheld device, the Sony CLIE.

Upon receipt of the grant in spring, 2002, our first order of business was to contract with the author of the KingKanji software to make the stroke order demonstration more viewable by displaying the character in a large pop-up box and at a much higher resolution. We also began looking into new educational software, including some exciting offerings such as the Walking Japanese-English Dictionary, Learn Japanese I, and the HandStory browser.

The CLIEs

Our research showed that the Sony CLIE (T665Cs and NR70 series) offered dramatic improvements over the Visor: higher resolution (320x320 instead of 160x120), color, a built-in MP3 Player, a built-in Movie Player, and a built-in Image Viewer. Best of all, it supported the Memory Stick, capable of storing 128MB of software, movies, images, and sounds. The capacity and ease of updating information on the Memory Stick may obviate the need for students to set up and regularly run the HotSync synchronization software, thus making it more realistic to access updated content on the handheld. In our view, its only weakness is the CLIE’s inability to record audio.

Pedagogical Possibilities

Throughout this project, we have maintained our core goals of increasing students’ long-term retention of linguistic material and providing tailored exercises for practice outside of the classroom. We have engaged students more actively in their own learning in a digital environment that seems helpful to most of them.

We are beginning to think of the handheld as a self-contained learning system that can provide not only complete portability (learning “anytime, anywhere”), but also “language-as-needed” rather than “all-the-language-you-need.” By providing a pared-down but essential reference system (Japanese-English and English-Japanese dictionary, a brief dictionary of grammar), a practice mechanism (character and vocabulary exercises, limited cloze, multiple choice and true/false exercises), listening materials accompanying the text, and some carefully chosen authentic cultural materials (audio, video, still photos, etc), students could access in the palms of their hands all the essential preparation for classes and tests. In concrete terms, we have already begun preparing MP3 versions of audio materials that accompany the textbook (with the permission of the publisher).

There are important implications in this technology beyond the language classroom. We are beginning to see the handheld as a device that would allow integration in some exciting ways. For example, the handheld could provide a way to link the content of language classes (Chinese and Japanese) with introductory Asian humanities courses. We could infuse linguistic content into cultural studies courses, and vice versa. Ideas and images could be linked across courses and even off-campus in the study abroad programs.

To better link our Japanese language courses with our Asian humanities courses, we intend to more closely tie language material with the Asian Studies theme-based materials. This will result in linking visual and audio resources with language tasks, allowing students to, for example, view clips of anime (Japanese cartoons) or manga (Japanese comics) that can tie into a theme.

The new version of the Sony CLIE, the NR70V, includes a built-in digital camera. We envision sending students abroad with these new units, enabling them to capture images (of signs, people they interview, etc.) and bring them back to ask questions about or to share as part of a presentation to fellow students.

For faculty members there are numerous possibilities:

1) keeping a bibliography (such as EndNote) at hand for research projects;

2) creating a database of one’s private library;

3) downloading and reading newspaper articles in target languages;

4) creating simple but interactive true/false and multiple/choice web-based reading comprehension and grammar exercises

5) with a portable keyboard, using the handheld as a laptop for word processing, spreadsheets, and PowerPoint presentations.

We also look forward to future handheld models that will allow us to record speech and send the resulting files electronically to the professor and/or other students. As wireless networks become more commonplace and are supported on the CLIEs either via 802.11b or Bluetooth, the amount of authentic language material that we can access quickly via a handheld web browser will also be powerful.

Conclusion

We met our initial goals, but as often happens in a project like this, we have ended by expanding our goals for using handhelds in the learning of Japanese. We also think that our experience with handhelds, preliminary though it may be, indicates some important implications for the learning of other languages and even for learning in other disciplines.