The “True” Corpus of American English

In an analysis of the Corpus of Historical American English (COHA) and Google Books, Professor Mark Davies of Brigham Young University studies the effectiveness of both engines and their ability to properly read the English language. Both are corpuses of the American English language, but Google Books has 500 billion words compared to COHA’s 400 million. Davies argues that although COHA has a significantly smaller database, the trending patterns still mirror those of Google Books. Because of this, Davies argues that COHA is actually the more effective corpus—a smaller database means far fewer data to sift through and that means quicker search results and faster information.

COHA’s “toys” are what make it a more useful database in Davies’s eyes. While Google provides the same basic function (showing the frequency of word usage throughout the decades), COHA is able to track concepts, related words, and changes in meaning. Whereas Google tends to have a one-track mind, just like it’s general search engine, COHA manages to “think” about relations for the words being placed into the search. Looking for things such as relation, form, root words, or even cultural shifts, the searches are much more comprehensive.

Design is a huge issue for some researchers, and with this in mind, Google definitely has the upper hand. True, as Davies puts it, COHA is able to effectively portray the same statistics, but I had a genuinely hard time navigating the site. Bar graphs are nice for portraying how many dark-haired, blue-eyed people are in a class, but for a corpus of American English, I found them rather ineffective. Google Books had a much more pleasing site, nicer to the eye, and easier to follow the pattern of the language. Yes, COHA’s tables are nice for their alternative searches, but as I said, traversing the site is actually rather difficult. Google provides a much more streamlined site, and provides actions that are easy to follow

Unfortunately for Google, I think the amount of words in the database have made it impossible to create proper analyses for the grammar, word meanings, and word foundations that COHA is successfully able to analyze. If Google Books created an efficient way to sift through all of that information quickly enough, then it would immediately become the preferred site. However, because of its inability to process the large amounts of data (largely its own fault) has rendered it ineffective to be a “true” corpus of the American English language.


Leave a Reply

Your email address will not be published. Required fields are marked *