Learner English on Computer

This is a collection of papers edited by the founder and coordinator of the International Corpus of Learner English (ICLE), which brings together written texts produced by non-native speakers (NNSs) of English from a variety of European mother-tongue backgrounds. The book is divided into three parts, each composed of several papers: the first part is devoted to a general outline of the constitution and analysis of learner corpora, the second to a series of case studies, most of which involve the comparison of NNS and native speaker (NS) corpus data, and the third to the use of corpora to enhance pedagogical tools such as dictionaries, instructional software, and teaching manuals. The opening paper of Part 1, by Sylviane Granger herself, is both an apology for and a description of the use of learner corpora. With links to previous work on error analysis, such corpora represent a sharp break with most current second-language acquisition research, which relies primarily on elicited data or introspection, and which often prefers oral to written data. Granger provides a good discussion of the design criteria that underpin the ICLE corpus, as well as the methodological issues involved in comparing NNS and NS data. In the second paper, Fanny Meunier provides a basic overview of the range of tools available for marking up raw corpora, from basic POS taggers to parsers to semantic, discourse, and error taggers. Also discussed are the types of software available for lexical frequency and concordance style analysis. The paper is complemented by a list of software in the appendix to the volume. The second part of the volume contains eight papers. H~kan Ringbom provides preliminary evidence of similarities and differences in lexical frequencies of advanced learner English in seven western-European learner corpora, as compared to a large native speaker corpus of argumentative essays. Gunter Lorenz discusses some stylistic aspects of adjective intensification in a corpus of German NNS, as compared to English NS data, finding among other things that the former tend to use intensification typically in thematic (clause-initial) position, whereas the latter tend to prefer rhematic (clause-final) position. Sylvia de Cock Sylviane Granger, Geoffrey Leech, and Tony McEnery present software designed to capture continuous formulaic expressions in NNS corpus data, and illustrate this using French speaker data, with special reference to vagueness tags such as or whatever. Bengt Altenberg and Marie Tapper discuss the overuse and underuse of adverbial connectors such as furthermore (overused) and however (underused) in a corpus of advanced Swedish learners of English, as compared to a corpus of native speaker writing. Tuija Virtanen looks at the frequency of