Analyzing a Topic.
Searching library databases can be difficult, but it doesn’t have to be. By spending a little thinking analytically about your topic, before you begin searching, you will save time and effort, and your searches will be more successful. This video shows you how.
Before we talk about library databases, let’s think for a moment about Google. When I use Google, I don’t have to give much thought to my search terms. That’s because web searches tend to be fairly simple:
- Academy Awards,
- New York Yankees,
- Flying Ants, and
- Elena Kagan
are examples of popular web searches1. For a typical web search, I don’t spend much time thinking about my search query: I just type the words I would naturally use to describe it. If I want information on flying ants, then I type “flying ants” into Google, and get results that meet my needs. With this search, I go from imagining the topic, to searching for it, with very little reflection on the words I use in my query. It can almost seem as if the search engine understands my question, and then answers with relevant documents.
The reality, however, is that when we use databases, we are searching for words, not topics. Most search engines are completely literal when interpreting a search query, and it’s important that you understand this when doing historical research.
If I search an article database like JSTOR for “native americans”, what I’m getting are not articles about the subject Native Americans, but articles in which the phrase “native americans” appears at least once. Articles about Native Americans that don’t use the phrase “native americans” will not turn up in my results.
There are over 7,000 history articles in JSTOR that use the phrase American Indians, but not the phrase Native Americans. Many of these articles are clearly about Native Americans, but were missed in my original search.
When doing historical research, it will often be the case that no single term is sufficient for finding all relevant articles. Therefore, before I search a database, my first task is to predict the words that would appear in articles on my topic. This is not always an easy task!
Let’s take another example: I want to find contemporary newspaper accounts of the court case Brown vs. the Board of Education. The phrase “Brown v. the Board of Education” doesn’t begin appearing in newspaper articles until May of 1954, when the Supreme Court announced its decision. If my only search term is “brown v. the board of education”, then my earliest article will be from May 19542. The case, however, began in March 1951. Does this mean there was no newspaper coverage of the case during the 3 years leading up to the Supreme Court decision? Not necessarily. If I use other search terms, like
- “brown” and “topeka”,
- “naacp” and “topeka”,
- “jim crow” and “topeka”, or
- “segregation” and “topeka”,
then I will find articles from as early as March, 1951.
The words I use will depend in part on the kind of documents I am searching. If I am searching for recent journal articles on Native Americans, then the two terms “native americans” and “american indians” might be adequate as a starting point. If, however, I am searching for articles in 19th century American magazines, then I would use different terms to represent my topic, terms like:
- “red man”,
- “aborigines”, and so forth.
Note that I sometimes have to use vocabulary that I myself would consider offensive, but this is the language used in publications produced at an earlier time.
Now, Native Americans would be an unusually broad topic for a history research paper, and it requires a fairly simple search technique: a single set of synonymous search terms to represent the topic. History research papers, however, are typically about more complex topics. For example,
- Immunization of Troops during World War II,
- Elopement in Early Modern Europe,
- Political Culture of Mount Lebanon in the late 19th Century,
- Heinrich Krippel’s Statues of Kemal Atatürk, and German Influence on Turkish Nationalism, or
- Native American Puberty Ceremonies.
These types of topics require more complex searches.
How to do Complex Searches?
Library databases are designed to handle complex searches. At first glance this search interface might appear cumbersome, especially when compared to the apparent simplicity of Google’s search interface; but unlike Google, this database is designed to handle complex searches. (Google actually does offer an advanced search option, but few searchers ever avail themselves of it3.)
If a simple search is for a topic like Native Americans, then a complex search might be for a topic like Native American Puberty Ceremonies.
Remember that, when I search in a database, I’m searching for words that might appear in articles on my topic. For this topic, I can begin with the words I already have, and type them as a phrase into the search box. Let’s use the database America: History and Life, which provides access to journal articles, dissertations, and books on the history of North America. The phrase “native american puberty ceremonies” retrieves zero articles. That does not mean there are no articles in this database on Native American Puberty Ceremonies. It only means that there are no articles with that particular phrase. So what other words should I use? You can see that with a topic like this, it’s not always possible to find synonymous terms, the way I was able to do with the topic Native Americans. In order to search a topic like Native American Puberty Ceremonies, I must first break it down into more manageable parts, or facets.
A facet is an aspect or a dimension of a topic, a conceptual part capable of being described separately with its own roughly-synonymous terms. So while I might be unable to think of synonyms for Native American Puberty Ceremonies, I should be able to do so for its facets.
For this topic, I have three facets:
- Native Americans,
- Puberty, and
These facets represent three parts of my topic. For each of these facets, I need to think of synonymous or related terms. For the first facet, I can use the terms: “native americans” and “american indians”. For my second facet, I will try the terms “puberty”, “adolescence”, “teenage”, “menstruation”, and “menarche”. For the third facet, I’ll use “ceremonies”, “rites”, and “rituals”.
This time, when I search using my new list of terms, I retrieve nine records. Whenever I do a search, I look at the records to see if there are any more search terms I could be using. One term that jumps out at me in all these records is Indians. Remember, the database is completely literal. The search term “american indians” does not include records where Native Americans are described simply as Indians. If I add the term “indians” to my search, then my results jump from nine to thirty-three. I’m already on the right track. My lists are never finished. Searching is a recursive process, and I will add new terms as I discover them.
This is how you analyze a topic. Analyzing a topic is the first step when preparing to search in library databases. The better you can describe your topic, the better the search results you will retrieve. Using these techniques will save you a lot of time and effort when doing historical research. In the next video, you will learn how to use the results of your analysis in an actual search.
1. These search examples are from Google Hot Topics.
2. This search was performed using proximity operators to compensate for the variations in the rendering of “Brown versus the Board of Education”. We searched for the word “brown” within 2 words of the word “board” within one word of the word “education”: brown PRE/2 board PRE/1 education in ProQuest Historical Newspapers.
3. Jillian R. Griffiths, and Peter Brophy, “Student Searching Behavior and the Web: Use of Academic Resources and Google,” Library Trends 53 (2005): 539-54; Jayadev H. Kadli, B. D Kumbar, and Satish Kanamadi, “Students Perspectives on Internet Usage: A Case Study,” Information Studies 16, no. 2 (April 2010): 121-30; T. Saravanan, P. Ushadevi, and V. Senthilkumar, “Google Use and Users: A Survey,” Information Studies 16, no. 1 (January 2010): 49-64.