We used the word trends widget for the terms “race” and “war.” We found that the words had a fairly consistent inverse relationship. “Race” begins as a commonly mentioned word in the first few issues, but declines as the use of the term “war” increases. Upon close reading, the reader would discover this trend derives from the US involvement in World War One. “War” reaches its peak use, unsurprisingly, in September 1918. This trend suggests that talk about race becomes less important relative to the war effort over time. From there, the trend reverses yet again after a few months, with the use of race becoming more frequent than war. With the First World War over and done with, race becomes a more relevant talking point again.
Another interesting find resulting from the search terms “race and “war” is the sudden lack of their use in May 1917. One would think that both would be mentioned at least once, either from historical context or even just by strictly looking at the trajectory of the word trends widget. Upon investigating, we found that the “issue” isn’t an issue at all, rather, it’s merely a supplement. The only subject covered is a horrific lynching. Given this context, it makes perfect sense that the words race and war aren’t used. Therefore, the supplement should not be allowed to skew our data. It is important to note that only through close reading can this information be ascertained.
To conclude, valuable information can be gained through reducing texts to graphical primitives. We were able to discover the inverse relationship between “war” and “race,” which could serve as a first step to close reading and a potential thesis. However, what is lost is valuable context. Without human interpretation, data can be misleading (and ultimately meaningless, but that’s a different point) like it was in the May 1917 supplement.