Approaching Language Transfer through Text Classification
Explorations in the Detection-based Approach
Edited by: Scott Jarvis, Scott A. Crossley
- Paperback - 200 pages
- 14 Mar 2012
- Multilingual Matters
- 234 x 156 (R8vo)
• This book gives the most comprehensive explanation of the detection-based approach to the investigation of crosslinguistic influence that has yet been published.
• This is the first collection of studies situated within the detection-based framework.
• The studies in this book represent the most comprehensive and in-depth examination to date of the nature and extent of crosslinguistic influence in the International Corpus of Learner English (ICLE).
Recent work has pointed to the need for a detection-based approach to transfer capable of discovering elusive crosslinguistic effects through the use of human judges and computer classifiers that can learn to predict learners’ language backgrounds based on their patterns of language use. This book addresses that need. It details the nature of the detection-based approach, discusses how this approach fits into the overall scope of transfer research, and discusses the few previous studies that have laid the groundwork for this approach. The core of the book consists of five empirical studies that use computer classifiers to detect the native-language affiliations of texts written by foreign language learners of English. The results highlight combinations of language features that are the most reliable predictors of learners’ language backgrounds.
In this bold and pioneering interdisciplinary study, experts on SLA research, computational analysis and statistics collaborate to try to identify the L1 background of non-native writers. The result is a most impressive work which will take the field of crosslinguistic studies a long way forward. A MUST for all SLA researchers!
Håkan Ringbom, Emeritus Professor, Åbo Akademi University, Finland
This is an articulate, comprehensive, and timely volume on a fascinating yet largely underexplored area. Jarvis and Crossley have produced an impressive collection of research-based evidence on language transfer using a corpus-based approach. The volume is a must-have for students, scholars, and practitioners interested in language transfer, corpus linguistics, forensic linguistics, text classification, second language writing, error analysis, and language assessment.
Pavel Trofimovich, Concordia University, Canada
This book indeed opens a new path in the study of language transfer. The investigations brought together here combine the strengths of earlier work on transfer with those of two other fields, computational linguistics and corpus research, which have much to offer students and researchers interested in second language acquisition and multilingualism. There can be little doubt that the tools used in this seminal work will continue to offer important insights for a long time to come.
Terence Odlin, Ohio State University, USA
Scott Jarvis is an Associate Professor in the Department of Linguistics at Ohio University. His work focuses on crosslinguistic influence and lexical diversity, with a special emphasis on methodological problems and solutions. Among his better-known works is the book 'Crosslinguistic influence in language and cognition' (Routledge) co-authored with Aneta Pavlenko.; Scott A. Crossley is an Assistant Professor at Georgia State University. His work involves the application of natural language processing theories and approaches for investigating second language acquisition, text readability, and writing proficiency. His current research interests include lexical proficiency, writing quality, and text coherence and processing.
Postgraduate Research / Professional