Defining Research Objectives for the Annotation Approach in an Oral English Learner Corpus

Plenty of research studies suggest the connection between L1 and L2 in that native language predetermines our ability to acquire various aspects of a foreign language. One might be more successful if the corresponding structure is present in L2 and, otherwise, encounter a difficulty if a phenomenon is completely new. Learner corpora serve as a vivid and convenient tool for language analysis allowing to reveal similarities and distinctive features between languages thus improving educational strategies. Hence, the objective of the work is to study answers from IELTS speaking module given by Russian native speakers and to create and elaborate an annotation system that would target common mistakes made by English learners, provide data on the frequency of different types of errors. An ultimate aim is an invention of an annotated learner corpus based on forenamed data. Spoken answers were studied and an annotation system of tags specifying error types was devised. As oral learner corpora tags differ from standard learner tags in that sentences are often incomplete, the tags were adjusted to the distinctive features of spoken language. Besides, tags that reflect students’ native language influences were added. The findings will facilitate the development of new educational strategies, namely, the efficiency of language comprehension is educed and existing study materials are revised. The results are also applicable in lexicography and in machine processing of language texts.