I got 40 topics after I uploaded the American South: North American Slave Narratives to Topic Modeling Tool. Below is a picture of topics from 0 to 26. Among those 40 topics, I chose NO.12 as my topic: bond, water, scott, carver, dr, cotton, sugar, put, add, 1, peanut, make, cut, sweet, gin, seed, plant, corn, potatoes, and milk. Since most of the items are related to food, so I made a hypothesis that this topic is about trading goods. In addition, I also found that “cotton” and “sugar” were products of slavery and slave labor, so I also assume that the topic could be related to slavery goods.
After making the hypothesis, I clicked on the topic and I found there are 204 documents in this topic. I decided to select top 10 documents to put into Voyant because these documents have higher number of word that assigned to this topic. And then here is the data generated from Voyant.
From the cloud, I selected Top 85 words that appears most frequently in the documents and I was surprised that most words has nothing to do with food, such as ‘time’ (appears 1578 times), ‘man’ (appears 1373 times), ‘mr’ (appears 1548 times), ‘said’ (appears 1467).However, these words seemed meaningless because it had nothing to do with my hypothesis, so I looked into words that has relatively less appearance but might relate to my hypothesis, such as ‘work’, ‘negro’, ‘white’, ‘school’, ‘money’, ‘life’ . Then, I selected some of these words and Voyant generated into a chart below.
From the chart, I didn’t find anything interesting except that the line ‘white’ and ‘negro’ rise and drop together. Then, I also looked into other parts of Voyant.
To conclude, by analyzing data from these documents, I realized that my topic is not about trading goods neither slave produced goods. My topic is about the life of slavery, for example, what they went through as a slave? What was the social environment at that times?