This paper presents the implementation of an action research whose main intent is to contribute to grammar teaching through contemporary means, also exploiting the cross-linguistic issues related to SLA (Second Language Acquisition) and ELT (English Language Teaching). This action research is comprised of three parts: The first part includes the detection of a problem; namely, errors in grammar, the second part of the research is based on the frequency of the errors compiled from the instruments produced as documents, and the third part includes a descriptive implementation of the corpus search in a small-scale Contextual Grammar course so that it can be modelled as a classroom procedure for the other teachers who wish to use novelty in their lessons. It is a fact that grammar is the foundation of all language forms, regardless of its being oral or written; however, how it is dealt with by different genres of language, styles and registers of each language specify the amount and kind of the grammar in the form of input to be used in the classroom. The study has introduced fruitful results in terms of the errors, and when corpus is applied as a new approach to Contextual Grammar teaching, it is sure to be of great benefit for the learners. However, teaching grammar using the corpus requires the fine selection of the grammar item, under the light of SLA to be raised as the course material. Teachers should be educated in initial teacher programs or trained after they start the profession for the efficient use of the corpora, for it needs fine tuning in terms of the selection and classification of the sampling before presented as the educational material in the classroom.
Corpus, contextual grammar, data driven learning, focus on form(s), foreign language education, inductive teaching