This paper discusses linguistic issues that one must address in order to design an e ective CALL system for second language learners. We focus on how one should develop an error taxonomy and indicate how that taxonomy can a ect the design of the entire system. This work was done in the context of designing a CALL tool to help native signers of American Sign Language (ASL) learn written English. Here we report the methodology used in developing the error taxonomy for the system. Our analysis of writing samples from ASL natives indicates that language transfer (LT) (when considered appropriately) can account for many of the errors we found. Recognizing this possibility has substantially in uenced the error taxonomywehavedeveloped. The resulting taxonomy captures the true source of many errors and will allow the eventual system to take advantage of ASL knowledge when teaching English. We provide a characterization of language transfer that provides a broad perspective on the ways that two languages may di er and indicates how these di erences could potentially in uence acquisition and production of the second language. In addition, we argue that an e ective CALL system must not only examine sentences in isolation, but it must identify and correct discourse-level errors. We point out that the source of several error classes we identi ed rests in discourse-level processing. We showhow these errors can

The ICICLE project is a Computer-Assisted Language Learning (CALL) environment geared toward teaching English as a second language. This paper reports on an initial prototype application of the system for deaf learners of written English. A primary focus of the ICICLE effort has been devoted to enabling the system to adapt to a learning user over the evolution of language proficiency. In this paper, we overview and motivate the design of our novel user modeling component that integrates Selinkers Interlanguage theory and other research in Second Language Acquisition in order to accurately represent a learners internal grammar as it changes over time. The objectives of this effort are two-fold: to accurately diagnose and respond to learner errors, and to focus tutorial feedback on those errors which are most relevant to the

Correcting text which is ill-formed with respect to grammar and/or discourse strategies is a challenging problem. We are working on this problem from the perspective of helping deaf writers produce text which conforms to the standard rules of English. 2 This perspective mayprove tobe particularly interesting since the native language of some deaf writers is American Sign Language

Abstract. This paper discusses the design and evaluation of an implemented user model in ICICLE, an instruction system for users writing in a second language. We show that in the task of disambiguating natural language parses, a blended model combining overlay techniques with user stereotyping representing typical linguistic acquisition sequences captures user individuality while supplementing incomplete information with stereotypic reasoning.