Need for data-driven instruction of engineering English

The author emphasizes the role of data-driven corpus-based methodology that promotes the study of natural language on examples of real life language use in teaching English for Specific Purposes (ESP), such as Engineering English. This approach can improve the way Engineering English is taught and add more creativity in ESP classes. In particular, the author demonstrates how vocabulary teaching in the ESP context can benefit from a corpus-based approach. She addresses the issue of technical and sub-/nontechnical vocabulary with characteristic examples from teaching situations in Southeast Asia. It is argued that sub-/nontechnical vocabulary should be given more attention in the ESP classrooms than technical vocabulary. Surprisingly, sub-/nontechnical words seem to cause more problems for ESP students specializing in technical fields than technical terms. It is also claimed that the availability of specialist language corpora and computerized text analysis (i.e., concordancing) programs can help learners acquire what D. Bolinger (1976) called "the prefabs of language," or formulaic multiword units (collocations), for technical and nontechnical uses.