Workshop: Introduction to Voyant Tools
Important Technical Notes for This Workshop
- This is not a required lesson for the course (unless your chosen Assignment Option requires it). Instead, it is a guide to the basics of using Voyant Tools for text analysis. NOTE: In Term 1 this Workshop is needed for success in completing Option A for the second assignment in Term 1 [2021-22]). As is the case with most parts of this course, it is best to work on this Workshop using a laptop or desktop computer rather than a tablet or smartphone.
- If you experience any problems viewing the interactive Voyant Tools windows, you should try to view the Workshop using a different web browser. For example, we find that the displays do not work effectively in Chrome, but they do work in Firefox. Many students report problems with Internet Explorer. Regardless of your browser type, check to make sure that you are using the latest version of it, since older versions of all browsers may result in problems using the interactive programs in Voyant Tools.
Big question
How can historians use software to help read textual sources?
Introduction
In past versions of the course (HIST 2F90) students worked together to transcribe several eighteenth-century texts about slavery. [NOTE: We don't do this in the 2021-22 version of the course.] The resulting digital files are valuable for historical research, because once these texts are in a digital format, they are machine-readable. This machine-readable format allows you (and other historians that we might share our files with) several powerful options for text analysis. Scholars sometimes use the term "distant reading" (or "text mining") to describe the techniques of analyzing sources with computer software.
In this Workshop you will learn about one program for distant reading: Voyant Tools. Two excellent ways to learn about these options are
- to read about and play with other researchers' projects online;
- and to build your own project.
Learning outcomes
At the end of this Workshop you should be able to:
- Experiment with distant reading by exploring a collection of many hundreds of runaway slave ads from the early United States;
- Create a shareable output of a Voyant Tools page that you upload in Sakai (we'll create a "Sandbox" Forum for you to experiment in)
Questions to consider, and learning activity
The learning activity for this Workshop consists of two main parts.
PART 1: Learning about Voyant Tools and Distant Reading
This part of the Workshop requires you to read Alyssa Anderson's blog on "Using Voyant Tools for Text Analysis." The essay introduces you to a Rice University project to read over 2,500 (!) runaway slave ads from the nineteenth-century United States. Of course, those are far too many ads for you to read in a conventional way, even though they're short. But what if there was a way for you to "read" these in a few minutes and actually have a chance to make some useful hypotheses about them?
Voyant Tools and other applications for distant reading allow just these kinds of possibilities. The interactive graphic below is a collection of several tools for analyzing ALL the ads at once. Notice that in the frame below you can click on individual elements of the program (or even hover over them with your mouse) to find out more about the slave ads.
After you read Anderson's essay, come back to this window, or better still go to the fuller page view of it that is available by clicking this link. See if you can notice any potentially significant patterns in these data. When you search for patterns this way, you will be starting to think about the significance of the sources. This kind of reading is not a replacement for close reading, but it can help you improve your close reading by helping to train you to ask questions of an entire collection of sources, or about the entirety of one large source.
You should also budget about 6-10 minutes to watch Tom Lynch's introduction to using Voyant Tools. Lynch is an education professor at Pace University. His introduction focuses on "reading" Herman Melville's very long nineteenth-century novel, Moby Dick. (If you are not familiar with Moby Dick, you might look briefly at the the Wikipedia page on the novel.)
NOTE: In Anderson's blog and Lynch's video an older version of Voyant Tools is on feature. You will learn to use the latest version of Voyant. It looks a little different than the original version, but the basic principles of its use are the same.
PART 2: Creating Your Own Voyant Tools File
In this Part of the Workshop you will create an exportable analysis of a Wikipedia page. Follow these instructions:
Make sure you are connected to the web. Voyant Tools can work off-line, but making that possible involves many steps.
- Choose a Wikipedia page from the Lesson on "Rulers and Subjects (I)" -- or another one of your choice. Longer pages are better than shorter ones.
- Open up the Voyant Tools page at http://voyant-tools.org/.
- Open your chosen Wikipedia page; then highlight and copy its contents.
- Make sure you only copy the main text of the article. In other words, be careful to avoid copying the bibliography, notes, or menu material in the left-hand margin. The reason is that this "tertiary" material will skew the results in Voyant. We don't want the program to "read" these extra data.
- Go to the Voyant window in your browser. Paste the article into the "Add Texts" window of Voyant.
- Click the "Reveal" button in the "Add Texts" window.
- Once Voyant has generated an initial analysis of your Wikipedia article, you will notice a word cloud (aka cirrus) in the top-left frame. When you move your cursor to the top-right of that small word cloud frame, several icons will appear. The first one on the left is a box with an arrow in it. Click on it.
- In the "Export" window that appears, choose the option "Export View (Tools and Data)".
- Then choose the option "an HTML snippet for embedding this view in another web page" ... and then click "Export". A window will pop up.
- Copy the text that appears in the window.
- Go to the Forum in Sakai and "Start a New Conversation".
- Add a title for your post (e.g., "Word cloud of Wikipedia entry on ..."), and then click on the "Source" option in the text frame (it should be on the top left). Once you've selected "Source," paste the details (i.e., the HTML snippet points 8 and 9) from Voyant there. Then click "post". This will embed your Voyant output in Sakai; if you do not see a Voyant word cloud display of your Wikipedia article, then you have missed something.
- Finally, ask yourself what the output from Voyant might mean or show -- or what the limits might be of the conclusions you can draw from the outputs. Instead of drawing quick conclusions from the outputs, focus instead on draw tentative hypotheses about the possible meanings of the outputs, and outline some initial questions about the text that you can explore further.
Try using Voyant to read other texts: for example, something that you have written, or perhaps a long novel that you find in digital format in Gutenberg.org (a repository of famous literature). Share your thoughts in the Sandbox Forum, if you wish.
Sources
- Alyssa Anderson, "Using Voyant Tools for Text Analysis," Digital History Methods, Rice University.
- A Wikipedia page of your choice from the Lesson on "Rulers and Subjects (I): Traditional Authority"
