Word Usage in Sacred Texts

During the 2016 presidential election, I spent a lot of time analyzing the word usage of speeches and debates and I thought it was interesting how much information you could glean by just analyzing the most commonly used words. As I was working on those projects, I began to wonder what I could learn by performing a similar analysis on the scriptures of various religious traditions. As I’ve noted previously on my posts, I’ve always been very interested in the world’s religions—their histories, traditions, belief systems, differences between them, and also their similarities. So much of this comes from each faith’s scriptures, so I thought that analyzing the word usage at a macro level could be particularly insightful.

So, in the fall, I started working on a series of blog posts and visualizations analyzing word usage in sacred texts. I started my analysis with the Bible (Judaism and Christianity), then moved onto the Bhagavad Gita (Hinduism), followed by the Tao Te Ching (Taoism). I had created visualizations for the Quran (Islam), the Dhammapada (Buddhism), and the Upanishads (another from Hinduism), but had so many other projects going on that I never quite got an opportunity to formally post them. Below are images of all of them. If you’d like to take a closer look, click on the image to see the full visualization. Note: I explain m methodology in my Bible blog post, so if you’re interested, you can read more here.

      

     

At the time I created these, I was still very new to Tableau and was still overusing word clouds and bubble charts, so these weren’t necessarily the best visualizations, but they were fun, interesting, and insightful.

After posting my Tao Te Ching blog on LinkedIn, a reader suggested that, after I was done with all the texts, I should do a comparison of all six together. I thought that this was a fantastic idea and, almost immediately, began working on it. My first idea was to give the viewer the ability to compare the top words in each text to the other texts, so I gave that a shot using sigmoid curves. Here’s what I created (again, clicking on the image will take you to the full visualization):

While I liked this idea and thought it looked really cool, it only allowed direct comparison of two texts. Sure, there are three sections to compare texts so you can kind of compare all six at once, but I felt that it still fell short of my goal. So, I decided to try something else. My second attempt was a simple packed bubble chart. Before you think it (or yell it at your screen), yes, I know the general consensus on packed bubbles, but with the density of my data set and without a need for exact precision, I thought that this worked fairly well to give an idea of word usage, particularly when enhanced by some highlight actions. Plus, I thought it was beautiful.

But, again, this didn’t quite give me what I was looking for. It was still difficult to see which words were most heavily used overall and how those compared across each of the texts, so I gave it one more shot:

Though not shown, this visualization has two tabs—one showing percent usage, the other showing raw usage. This version allowed you to see the overall most commonly used words then compare across all six of the texts. You can also sort by each text in order to get a slightly different view. After all was said and done, I felt that this version was probably most successful in telling the story.

This was about six months ago, at a time when I had created a lot of content and had a bit of a backlog of blogs and visualizations to post, so I added it to the list, expecting to eventually write a blog about it. Over time, my backlog dwindled, but I never posted the visualization. There was just something about it that felt unfinished. It seemed that none of the three properly told the whole story. So, after it sat on the shelf for half a year, I picked it up a few weeks ago and decided to give it one more shot. This time, I’d combine the elements of all three visualizations into a single dashboard. When it was done, I felt I had finally nailed it. So here’s the final version:

You may have seen this before since it was named Viz of the Day on Friday, July 21 (interestingly enough, I wasn’t even finished with it at the time—I scrambled to knock out the few remaining pieces that evening). As if that weren’t exciting enough, it then became Viz of the Week!! Anyway, I hope you enjoy the visualization and if you have any comments or questions, please let me know.

Update 10/03/2017: I just found out that this visualization had made it to the long list of the Information is Beautiful awards!! The short list will be announced on October 17. 

    

Ken Flerlage, July 28, 2017.

Twitter | LinkedIn | Tableau Public