(Cross-posted on the Google Research Blog)

It can’t have been very long after people started writing that they started to organize and comment on what was written. Look at the 10th century Venetus A manuscript, which contains scholia written fifteen centuries earlier about texts written five centuries before that. Almost since computers were invented, people have envisioned using them to expose the interconnections of the world’s knowledge. That vision is finally becoming real with the flowering of the web, but in a notably limited way: very little of the world’s culture predating the web is accessible online. Much of that information is available only in printed books.

A wide range of digitization efforts have been pursued with increasing success over the past decade. We’re proud of our own Google Books digitization effort, having scanned over 12 million books in more than 400 languages, comprising over five billion pages and two trillion words. But digitization is just the starting point: it will take a vast amount of work by scholars and computer scientists to analyze these digitized texts. In particular, humanities scholars are starting to apply quantitative research techniques for answering questions that require examining thousands or millions of books. This style of research complements the methods of many contemporary humanities scholars, who have individually achieved great insights through in-depth reading and painstaking analysis of dozens or hundreds of texts. We believe both approaches have merit, and that each is good for answering different types of questions...