John Mulaney Teaches Us Computer-Assisted Analysis

If you’ve never heard of John Mulaney, let me be the first to introduce you to a loveable, gawky stand-up comedian who got his start on SNL. One of his most well-known bits is about grade-school assemblies and the “street smarts” he learned as a child. He describes the colorful, real-life character of J.J. Bittenbinder, who “looked like he should be the conductor on a locomotive powered by confetti” and had a huge handlebar mustache. Bittenbinder’s assembly taught elementary students about stranger danger and tactical ways to escape from criminals.

Look at the man shimmy!

So what can John Mulaney teach us? With an excerpt from his act, I can show you how to make computer-assisted text analysis work for you. And it’ll be more fun, because it’s comedy! I’ll be using Voyant Tools to demonstrate how the site can manipulate digital text to show you themes and statistical relationships in the writing itself: https://voyant-tools.org/

Voyant Tools is a web-based reading and analysis environment for digital texts.

On the homepage you can insert text directly into the box, upload a file, or simply paste URLs and hit the “Reveal” button. Once you do, the site will create a list of all words used and track their frequency, mark their relationships, and provide context for the use of specific words. This information will be presented in multiple areas, oriented around the central text listed in the center box. In the top left-hand corner, a word cloud will appear that can be manipulated to show the most used words and their relationships to each other. In the right-hand corner, a list of all words used and their frequency will allow you to click on a word and see it highlighted in the central text.

While this may seem overwhelming at first, I would suggest playing around with a small piece of text and looking at the various trends, it gets easier to use with a little bit of practice.

On the bottom left-hand corner, you can view a summary of vocabulary usage and average sentence length, manage multiple documents, and view the most commonly used phrases. On the bottom right-hand corner, you can view word relationships, correlations, and context. Voyant Tools is a free resource that can be extremely helpful in analyzing historical sources and other forms of written text. Similar tools have been used to aid historians in condensing large blocks of text and sources into themed sections (see Cameron Blevins blog post about topic modeling and Martha Ballard’s diary here: http://www.cameronblevins.org/posts/topic-modeling-martha-ballards-diary/).

Some other potential uses of this site could include:

  1. Using the tool on oral history transcripts to track themes and narrow down the central parts of the interview. Oral history interviews are often not linear, and using a text analysis tool could aid scholars in organizing the narrator’s information, memories and insight.
  2. Using the tool on scholarly articles to aid in summary, understanding and relationships between argumentative points.
  3. Using the tool on historical documents that have been digitized, such as diaries, letters, unpublished manuscripts, etc. to support and enhance analysis.
  4. If you have any other ideas, drop them in the comments!

Before I go, I’ll just say a big thank you to John Mulaney for showing us how to use Voyant Tools, and if you haven’t watched his Street Smarts bit, it’s in the first twenty minutes of Kid Gorgeous at Radio City Music Hall (stand-up act available on Netflix). He also talks about ghosts and Donald Trump so it’s a real winner. If you have any questions or would like more insight into Voyant Tools, let me know! I am happy to answer any questions. I would highly recommend playing around on the site and learning how it works for yourself. It’s easy and can be fun, especially when you use entertaining chunks of text. If you want to use John Mulaney again, or pull from a movie or TV show you really like, transcripts can be found here: https://scrapsfromtheloft.com/

Happy analyzing!

8 Replies to “John Mulaney Teaches Us Computer-Assisted Analysis”

  1. Shae, I really liked your suggestion about using this as a tool for an Oral History transcript. I think most of us know first hand how frustrating doing transcripts can be and this is a great tool for organizing the information. This is a great suggestion for how to use it Shae and can also identify themes across multiple interviews to make easy work of a very time consuming task.

    1. Also, I wanted to ask does this work in other languages as well? Can it aid in translation? I can think of ways it would help someone translate but does the software do it on its own?

      1. Amanda, what a great question! I just tried it on a Spanish poem. I don’t see any tools for translation, but the rest of the analysis processes work the same way. The only glitch I could find was the correlations button, which seems to be taking much longer to make connections between the words (potentially because it isn’t in English, so grammatical correlations are different). So there might be a bit of an issue there, but otherwise, everything works the same!

  2. Shae, I think all three of your suggestions are really great ideas for using Voyant. I can see using the platform for any of the three ideas you said, but out of the three, I think that the idea to use it for analyzing historical documents makes so much sense. This would allow an in-depth review of a primary source and could reveal certain themes or important words/phrases in the document that one might not have realized in a first glance or a skim. This kind of idea is just one example of a tool that I was not aware of even as a graduate student, but I think it could be useful for even undergraduate history students, grad history students, or people in similar fields, especially as they are learning how to navigate primary sources and build skills relevant to the field. Also, I am so happy that you used John Mulaney’s J.J. Bittenbinder bit for your post because he is my all-time favorite comedian and Kid Gorgeous is a fantastic special!

    1. I completely agree with you about use for undergraduate students or even other researchers in the humanities (imagine how useful it would be for literary analysis!). I love Kid Gorgeous, and I’m glad you appreciated the use of Bittenbinder. Every time I talk on the phone with my father he reminds me of it by yelling “Street Smarts!” so I remember to stay safe.

  3. This is a great overview, and I loved that you referenced John Mulaney. I definitely found the pieces talking about this kind of digital text analysis to be very interesting. It’s fitting that you brought up oral history transcripts, since the transcript from my project last semester was the first thing I plugged into Voyant, and it worked very well!
    It also made me think of one of the readings we had last week, that was talking about the role of historians in a more open historical environment. The data you can get from these kinds of tools is great, but the analysis and comparisons that a historian can make is equally important to make that data matter. Basically, these kinds of tools can give us the “what” of something, but historians are needed to provide the “so what?” I think these tools are an important part of developing history as a field, and I’m excited to see what other tools modern historians can use.

    1. Shaan, I love that you put in your transcript! What connections did it help you make? How would you have used the tool last semester to speed up or change your analysis process?

  4. Shae, I really like this write-up, especially your suggestions on how this site might be useful for historians. Voyant Tools reminds me of Wordclouds.com, but with a better user interface, and more tools for text analysis. Your suggestion about using it for diaries and letters is very intriguing to me, because I have transcribed many letters, and although I can usually pick up on re-occurring themes, it would be useful to have a tool that can assist me in that analysis.

Leave a Reply

Your email address will not be published.