I recently decided to resurrect an old assignment that used the now-defunct Google News Archive search. This assignment was greatly successful because it not only data-mined historical newspapers for search terms, but students could easily locate sources to provide examples for how different terminology appeared over the years.
Here is the original assignment (trimmed):
In this paper, it will be your job to make use of the Google News Archive search (http://news.google.com/archivesearch). …After selecting your term and generating a search and a timeline, next you must then explain and contextualize the timeline in a paper of 3-4 pages (approx. 1000-1200 words). You have a number of choices here:
1.) You may choose to explain the timeline’s change over time. For example, you might identify two or three “peaks” of the occurrence of the term you chose, and construct an argument for why the term peaked in these two eras, explaining what differences occurred between the two ￼￼￼periods. What was different about each context? What issues remained the same, or unresolved?
2.) You may also choose to explain a single moment on the graph. Perhaps the peak of a certain term represents an important historical moment, where an idea or movement had special resonance.
3.) You can also be creative — perhaps there is another approach you feel is useful to explain the timeline you have discovered.
Whatever approach you choose, you MUST MAKE AN ARGUMENT It is not enough to DESCRIBE the timeline, you must also EXPLAIN WHAT IS SIGNIFICANT about it. What is going on behind the numbers? Why did people care? What changes and continuities are significant? How do historical discussion of familiar terms differ from present discourses? In making your argument (in the body of your paper), select two or three articles to help demonstrate your case. Your only sources should be the articles themselves (which will serve as your PRIMARY SOURCES), and sources from class (lecture notes, textbooks, other assigned readings). It is not necessary for you to conduct other research: I’m most interested in seeing how you contextualize these historical documents and debates, and demonstrate change/continuity over time. Remember: you have license to be creative with this assignment, but your paper MUST HAVE A THESIS.
Some terms to help get you thinking:
Labor; Anarchist; Suffrage; Liberal; Conservative; Nuclear; détente; Slum; Civil Rights; Sunbelt; White Flight; Black Power; Chicano; Feminism; Abortion; Religious Right; Debt; Income Tax; Suburbs; Immigration; Socialism; Negro/African American; Segregation/desegregation; Progressive
The assignment yielded some interesting papers. One student looked at the word “hippy,” while another examined “ghetto.” All the papers included the graph of their results, and other than the papers that were merely descriptive (i.e. they described the graph and the thesis essentially explained that the term increased in usage, without explaining why), I received some creative responses. The “ghetto” paper, for example, traced how the word shifted from international sources in the 1930s to describe America’s urban crisis in the 1960s and 1970s. Generally speaking, the old Google News Archive Search provided a useful way to show students that language is not static, and that terms we use on a regular basis are shaped by historical events and understandings.
So I recently found a post at edwired.org reminding me of the Google Ngram tool that rolled out before the demise of Google News Archive Search, which has vastly more potential (and a much easier name). News that the data behind the search has improved made me look again. While my students won’t be examining the intricacies of grammar, they might be intrigued by the ability to overlap word usage in one graph, which could help produce more complex and nuanced arguments.
I tried this with a few searches based on popular ones from student papers:
hippy/hippie and beatnik:
Even on their own each of these graphs could produce a worthwhile student paper. It might be worth including certain contextual terms like “juvenile delinquent” or other Cold War culture related terminology to help students think about the context in which these word shifts occurred when it came to youth rebellion. In the second graph, certainly some history of these words would have to accompany the graphs because students wouldn’t have examples readily available in the search results to contextualize the data for them, as newspaper articles helped with in the first instance.
I have a couple of concerns. One is that the results are divorced from the data; it would be difficult for students to envision the sources, and the variety of sources, from which the data derived. Since they would not be engaging in linguistics or the kinds of text mining and sophisticated results produced by digital miners like Dan Cohen, I’m considering limiting the number of possible phrases and search terms so as to provide a more guided experience — it would cut down on the creative responses, but might serve to elevate the products received.
Another concern I have is the multiple possibilities of such an inquiry. I’m thinking that with so many possibilities at their fingertips, and without the documents themselves to consult, students would get too lost in searching for possible terms. The solution to this, I’m thinking, will be to set up their inquiry more as a guided research project where I provide a few search options, plus some supplementary primary and secondary sources for them to consult.
So the question is, what would be some useful terms to search for in a class focusing on the United States 1945 – present? Does anyone have experience with successful searches that yield interesting results?