Words Lesson Plan 1- 1101

An introduction to using computers to analyze large amounts of text, or distant reading.

We are using a website that processes large quantities of data. If all students use the website at once, it will crash.

If you are comfortable installing software on your computer and have the rights to do so, directions for how to install Voyant are here.

Learning objectives:

Student will be able to:

  1. Enter URLs or text into Voyant.

  2. Demonstrate an understanding of the Voyant tool.

  3. Use stopwords.

  4. Demonstrate that they can draw conclusions based on their use of Voyant.

1. Distant vs. close reading. A lesson using historical religious and popular culture texts.

Most reading we do is “close reading.” We read each word, place each word in the sentence or context, and then create meaning out of the words all strung together. For example, “today, I ate cake.” You must read those words in context and in an order to undertand those words.

Sometimes, we read in ways that aren’t so “close.” For example, if you go to a weather website, and look up the forecast, you don’t read all the words in their context. You scan for the information you need, and ignore the rest. This is the first step to distant reading: recognizing that not all information included in a text is relevant and looking only for the material (or data) that is important.

For part of the Words module, we’re going to use distant reading websites to analyze large amounts of text. For example, below I’ve place a URL of the entire text of the Mahabarata, the other great Indian epic (along with the Ramayanal) into Voyant. This tool counts words and looks for patterns. It is almost impossible to count words in large numbers for multiple books as a human, but computers can do it for us. This is what “distant reading” means: humans are away from the texts and computers “hold” and manipulate the texts.

Distant vs. close reading. A lesson using historical religious and popular culture texts.
Distant vs. close reading. A lesson using historical religious and popular culture texts.

2. The entire Mahabarata, by its words and numbers.

I clicked “Reveal” and Voyant has now analyzed the entire text, and counted every word, generating what we call a word cloud. A word cloud shows the words used most often in a text. More popular words or symbols are bigger. in the word cloud below, “said,” “great,” “continued,” and “like” are the most popular, which is useless to us. So, we need to tell Voyant to edit out those common words. We call common words we don’t want “stop words.”

The entire Mahabarata, by its words and numbers.
The entire Mahabarata, by its words and numbers.

3. Editing stopwords

I clicked on the switch icon below the word cloud and it will give me the option to edd Stopwords.

Editing stopwords
Editing stopwords

4. Select Edit List and make sure “Apply Globally” is checked

You can also manually add words for voyant to ignore by clicking on “Edit List.”

Select Edit List and make sure "Apply Globally" is checked
Select Edit List and make sure "Apply Globally" is checked

5. Add any words to the list that you don’t want.

Here I’ve added “said,” “unto,” “hath,” and “continued,” and several other common words, hitting the return key after each word and then hitting “Save” and then “Confirm.”

Add any words to the list that you don't want.
Add any words to the list that you don't want.

6. New word cloud revealed.

New words emerge as

Clicking on any slider icon in the upper right gives you more options.

New word cloud revealed.
New word cloud revealed.

7. Clicking on icons gives you acces to more options.

Clicking on slider icon reveals more tools.

Clicking on arrow in a box icon reveals more of that section of Voyant.

Clicking on icons gives you acces to more options.
Clicking on icons gives you acces to more options.

8. What matters more: son, daughters, wife, or husband.

I added the words daughter, son, husband, and wife to the trends tool on the right (1).

Voyant split the Mahabarata into 10 segments, and shows the number of times each word is used in each part of the ancient Indian epic.

Looking at this graph, I might be tempted to argue that this text is most concerned with sons. Still, I would need to know more before I offered that argument. Voyant just counts words and shows those counts. It doesn’t offer explanations for anything.

What matters more: son, daughters, wife, or husband.
What matters more: son, daughters, wife, or husband.

9. It works for non-history too.

You can run other websites or text through voyant too. Try the below article and see what you find: https://www.citylab.com/equity/2018/10/searching-soul-community/572986/

The longer the article the better, because Voyant is best at counting words, and short articles can be read without any computers, easily.

It works for non-history too.
It works for non-history too.

10. How can distant reading help us understand the past?

As humans we can only read a limited number of words at a time. Computers in the form of software, however, can “read” huge numbers of words, entire libraries in fact, fairly easily. More importantly, software can count, compare, and display patterns in ways we can’t.

Computers also help by revealing patterns that ore contrary to our assumptions, which is especially important when studying religion.

For this week, I want you to get used to using Voyant with low-stakes texts with which you are familiar.

Assignment

  1. Run two articles of at least 1000 words through Voyant. The articles can be about any subject that is appropriate for the workplace and can come from any publication
  2. In complete sentences, what conclusions could your draw from Voyant? What do the word counts tell you about what the author valued or the people in the article valued? You should produce two paragraphs of around 150 words.
  3. Post the URLs of your articles in your document above your analysis.
  4. Save your document and post it as to our Experiment discussion board.

Example

For example, I could have Voyant analyze articles about a recent political or cultural event. After filtering for the names in the article, I would look for what places are mentioned most and what adjectives get used to describe people, places or things. The longer the article, the better the tool works as we can see patterns in words with shorter works without computers.

Grading Criteria

Student

  1. Entered URLs or text into Voyant.

  2. Demonstrated an understanding of the Voyant tool.

  3. Used stopwords.

  4. Demonstrate that s/he can draw conclusions about the values of the author or covered people based on his/her use of Voyant by writing two paragraphs.