Google Scholar and COCA-Academic: Two very different approaches to examining academic English

In a recent article in the Journal of English for Academic Purposes, Brezina (2012) compares Google Scholar to the 91 million word academic component of the Corpus of Contemporary American English (COCA). In this article, I examine this comparison and show that – with the searches done correctly – COCA offers much more data than Brezina suggests. More importantly, I discuss at some length the many types of searches related to academic English which are possible with COCA but not Google Scholar, including searching for constructions (using part of speech and lemmas), comparisons between academic and non-academic genres or between different sub-genres of academic, creating frequency lists, finding collocates (to examine word meaning and usage), and carrying out semantically-oriented searches with synonyms and customized lists. Finally, I show how the new WordAndPhrase.info site provides even more user-friendly access to COCA data, including the ability to browse through large frequency lists of academic English and input and analyze entire texts. All of these COCA-based searches provide a wealth of information for teachers and learners of academic English, and while they can be done quickly and easily with COCA, all of them would be difficult or impossible in Google Scholar.

► Overview of the academic portion of the Corpus of Contemporary American English (COCA). ► Responds to Brezina's (2012) claim that the 91 million words of COCA-Academic are too small. ► Shows ways COCA can examine academic English that are impossible with Google Scholar. ► These include searching by part of speech, and comparisons between ±academic and sub-genres. ► Also includes semantically-oriented searches via collocates and integrated thesaurus.