After doing a deep dive into the text on 3 specific sources I found that the most frequent words used are “Negro”/”Negros”, “White”/”Whites”, “Colored”, “Riots”/”Rioting” and”Police”. This goes to show an emphasis across the board on the racial and rioting elements of the event.
Barring racial terms the most common terms are those pertaining to locations and objects of attack such as cars and busses. I’m not sure if there was an issue with my data set but when I did read through the articles I saw many OCR errors both misspelling errors and completely wrong outputs consisting of random symbols. I was looking for more specific terms such as “Said” “Spoken” or “Mentioned” to see if I could determine which voices were elevated but I could not find anything of the like. The words I did find were not the most helpful. This may be an issue with my data pruning process though I was very careful not to prune off anything I could see as even the slightest bit insightful. The terms I did prune often were references to locations or terms like “island” “park” and “Isle” which are really just proxies for “Belle” which is already inside the dataset.
Comparing the newspapers against the report when we look at the distinctive use of words between them, the report favours the term “Colored” over the term “Negro” as it has the highest usage of the word “Colored” and the lowest for “Negro” despite it being the longest of the three sources. In general, it seems like the report attempts to use more clinical language trying to focus more on events as they happened and not a sensational account although upon reading the report in its entirety it is quite biased on where it places the blame. The papers use more colourful language such as “Beaten”, “Yelled” “Riot” and “Mobs” in larger numbers relative to how much shorter the news articles actually are. Both the Black and the White newspapers presented very one-sided accounts of the events and it’s unclear how much is rumour or reality.
I will say the white article starts off a bit stronger being more fact-based whereas the Black article does start with a more sensational headline and introductory story. But at the same time, I believe the Black paper was completely in the right to highlight the stories that it did because their voices were being drowned out everywhere else so even if these articles are a little sensational I think there is some value in the Black article that helped me get more context and learn about some additional stories I may have otherwise missed. As for the text itself, the Black paper favoured the word Colored as opposed to “Negro” and the opposite is true for the white article. There is more race-baiting in the white article when I opened up and viewed the context and I would say the preference for the term “Negro” suggests loaded language in some sense. Examples of this include the use of the phrases “Negro mob” and “Gang of Negros”. I’d also point out that words appear to take on different meanings between the articles. For instance, the word “Mob” typically implied white mobs in the black article and black mobs in the white article.
The Text Analysis reaffirms my decision to label this event as a Race Riot as all the keywords, the context, and even the relation between many of these words suggest a racially motivated riot. But I’d also say some good additional insights came out of this analysis for instance my assumption coming into this was that the term “Boys” was largely used as a racially charged reference to black men but when we view the text’s relational links “Boys” in these articles is used pretty evenly to describe both Blacks and whites alike with the term “Men” actually used more commonly based on this analysis in reference to Black men as opposed to whites. Even when I drill down and look into the context it appears not to be as loaded as my initial bias would inform me to believe.
