Mariëtte Dijkstra (2005)
GironiMOO: a Research Tool for Capturing Implicit Knowledge
Master's thesis, Rijksuniversiteit Groningen.
[ Paper (PDF, 952 kb) ]

Chapter 1: Introduction

Though opinions differ on what are the exact characteristics of written and of spoken English, it is usually presumed that there is a simple dichotomy between the two (Hughes, Freeborn). Written English is, in general, considered to be more formal than spoken English, to use a wider variety of sentence constructions and a more varied vocabulary and not to be used for real time communication (Hughes). This rather neat distinction has been blurred somewhat by the arrival of various electronic techniques for communication, such as Internet Relay Chat, MUDs/MOOs and instant messaging (IM), all of which use typed text on a computer screen to hold 'conversations' in (almost) real time. In practice, this means that the English used in these media, though in written form, takes on characteristics of spoken English (Johnson, Segerstad).

On a more abstract level, spoken language and spontaneous language production can be associated with implicit linguistic knowledge, as I will argue in this dissertation. Capturing implicit knowledge has proved very problematic to date, since it is impossible to access this type of knowledge directly and the speaker is unaware of it. Since spoken conversation is very hard to monitor, testing for implicit knowledge is generally done in written form, usually by having speakers answer questions about their performance. This type of testing is not very accurate as it can easily be contaminated by explicit knowledge, which I claim can be associated primarily with written language.

In this dissertation, I claim that because of its unique ability to capture spontaneous speech production in a form that can be monitored, the medium of chat could prove successful as a tool for eliciting and testing implicit knowledge. I will propose an electronic learning environment for beginning EFL learners, based on an educational MOO, which features chat between learners as its main method of communication. If it is indeed possible to capture the learners' use of implicit knowledge in it, this environment could prove very useful for future research in this field.

Chapter 2 provides a brief overview of the development of Computer Assisted Language Learning (CALL) and the two major underlying theories of learning. A sketch of the development of IRC/chat and of MOOs is given and the position of GironiMOO (as the proposed learning environment is called) within this framework is explained. In this chapter I will also discuss why English as it is used in the specific Computer-Mediated Communication (CMC) application of chatting and instant messaging can be described as written conversation and how it fits into the written/spoken language dichotomy.

There are many theories on second language learning, which?different though they may be?all distinguish between implicit and explicit knowledge and learning, where explicit knowledge is generally tied to written language and implicit knowledge is generally associated with spoken (spontaneous) language. In Chapter 3 I will discuss the differences between implicit and explicit knowledge and learning and how these tie in with written, spoken and, in particular, chat and IM English. The focus of this chapter will be largely on implicit knowledge and methods of testing and elicitation of this type of knowledge, since this is what I am interested in capturing in the electronic learning environment that I propose. Finally, I will discuss how this learning environment can contribute to current research in this field.

Chat as used in the learning environment I propose could prove successful as a tool for testing and eliciting implicit knowledge, since it captures spontaneous speech production in a form that can be monitored. Chapter 4 will deal with the details of the learning environment that are relevant for this type of research. The environment is based on an educational MOO and basically consists of the possibility for students to chat with each other in a (virtual) natural language setting, while also completing a secondary task. The chat sessions that take place between learners will be logged so that they can be analysed. The technical realisation of GironiMOO is also presented in this chapter.

The usefulness of this environment for research of implicit knowledge will be assessed based on a pilot study among a small number of beginning EFL learners. The complete experiment can be found in Chapter 5.

The final chapters of this dissertation will obviously be the conclusion in which I will discuss my findings, followed by some suggestions for future research.

This dissertation is part of a joint project between the Departments of English and of Humanities Computing. To reflect the different perspectives suggested by each department the project has resulted in two dissertations, one for each department. However, as a result of the joint character of the project, there is a considerable amount of overlap between the two.