This afternoon I played for a while with the google labs books ngrams program. This is still an experimental concept. The idea is to identify specific words and phrases from all books that Google has so far digitized and placed in specific datasets called “1-gram” units representing words.
Using these datasets it is possible to see how specific words and concepts evolved over time and what people are writing about. It is quite interesting to see how concepts like evidence have grown over the years particularly since the 1960s and how words like eminence that was once popular at the turn of the nineteenth and twentieth century has nosedived since the seventies. (To see what I mean, type the two words separated by comma in the ngram website and see the graph, quite interesting). This may have some interesting connotations if you start thinking of how societies have evolved in their thinking and practices and concepts over time.
Quite an interesting project to keep an eye on.
Can it change the way one conducts research? I think it does and that aspect interests me. In a flash one can pan generations of books and identify key words over time and how these words and concepts and ideas evolved over time and which ones were dominant, which went out of fashion and so on. If past is a window to the present, then a web service such as this is a great way to reivew and in some situations to teach research data analysis and graphics to students.