Instead of doing a close reading of over five hundred books dealing with the immigrant experience in the 19th-century to learn about specific themes, you can use a database to do it for you. Sounds like a time-saver, right? Yet this methodology is not widely used by historians in the field, but why? These are some of the questions that Matthew Jockers attempts to answer in his book Macroanalysis: Digital Methods and Literary History.
The practice of macroanalysis is when a scholar does a “distant reading” of a large number of books at the same time through a database like Google Ngram. The example that Jockers gives in his book is when he researched Irish American literature in the western United States. He took a set of about eight hundred books and was able to make a searchable bibliography. Jocker was able to chronicle Irish American fiction by reign, gender, and setting within his database. The Western Institute of Irish Studies used his database for some years as a research tool for others to use. This method has revolutionized the way researchers can attain information and how to present it.
Macroanalysis allows researchers to chart different things from a text. For example, the database can determine how many times the word “and” or “for” comes up in a sample of documents. Knowing this information is essential for those trying to analyze a text to figure out the similarities between writers. These databases can also create word clouds that can highlight what topics are more prevalent with a sample of books. Jocker praises the benefits of this methodology but warns researchers about the drawbacks.
One of the main concerns that come up is the fact that this methodology is not always accurate. The database can confuse categories together because they can seem similar to each other. As a result, informational graphs and tables can be misleading to scholars that wish to expand on the work. Along with this, a database has a hard time picking up on subtle phrasing that a human could.
Another drawback is copyright laws that prevent researchers from accessing literature. Jockers notes that when talking to a group of students, he “watched this same group move from excitement to despair” (175). This feeling came as a result of the students not being able to go passed copyright laws that force them to pay to use the material. Historians also have to face similar barriers when researching things like government documents that are under red tape. These drawbacks then beg the question, should this method be used?
The macroanalysis method is not a replacement for close reading analysis. A computer can miss many things that a person can pick up on like a double meaning of a passage. However, this is not to say that this methodology should not be used at all. On the contrary, this technique should serve as a research aid for those trying to dig through a large number of books. Perhaps one day we can have a database sophisticated enough to compute the most accurate information, but until then, this is another tool in our research kit.