Comparing Ngram tools
These pictures are souvenirs of word frequency visualization tools. These graph visualizations are the result of entering the words “war” and “love” in both bookworm and Ngram viewer. The bookworm tool shows that “war” is usually less frequently used in American literature as opposed to the word “love.” However, there are clear spikes of the word “war” during the Civil War and World War I. It only graphs from 1820 to 1920, whereas the Ngram viewer tool graphs from 1800-2000. The Ngram tool is therefore of greater use on a larger scale. It shows that “love” and “war” are being used at the same frequency in texts in 2000, which is the closest to modern day data. These differences are important because it shows that bookworm is better for a narrow research of 100 years, whereas Ngram viewer can show a broader overview of the terms frequency in different time periods. Ngram and bookworm encourages assumptions more than it values interpretation because the observations in this visualization do not offer any information beyond the year that these terms were frequently or infrequently in text. Based on this visualization, a person can only identify when the words are used most frequently without being about to think about what texts utilize both words or why these terms were important in different time periods beyond what the data presents.
This souvenir shows the visualization of one topic of the literature sample that I uploaded to Lexos. This tool uses the size and colors of words to display frequency of use rather than lines on a graph. This kind of visualization quickly shows the reader which words appear most frequently in a topic. Drucker argues that visualization tools in the digital humanties need to be reworked to be interpretive rather than to encourage assumptions. Word clouds are more humanistic than some visualization forms because it does not visualize time, but Drucker and Klein would encourage it to be re-worked. The immediate effect of the most frequently used words being bigger is that a viewer makes assumptions about the text. A way to re-work this visualization could be to make the words all the same size and continue to have the scroll-over curser reveal how many times they were used in the topic. This will remove the inclination to make assumptions based off size and color.
This visualization represents the “ideal bookshelves” of various people that were interviewed. The tool allows for someone to hover over a node and see what books people have in common on their shelves. The second souvenir shows that Rivka Gelchen has books in common with four different people, and of those four people, three of them put Catch-22 on their shelves. This network visualization gives less information than the graphs farther down the page, but is still important to analyze. It shows which people had the most intersections of books. Once that is known, the graphs farther down the page show the distribution of the professions of those interviewed. This visualization is factual, but I think that Drucker and Klein would agree that it allows for interpretation. The information is more of a tool to make connections than a presentation of assumptions.
This pie chart depicts where the greatest number of artists are born. More than half of them are born in the UK and 11.9% are born in America, which shows that most famous artists come from English speaking countries. After exploring the Google Fusion tool more, I think that the pie chart and the configure network graph could both be helpful in the final project. No matter which corpus we decide to use, the pie chart is a great way to visualize data about the authors. The configure network graph can show the connections between authors, topics, and literary works in general. We can use a variety of information to make connections in this way. Drucker and Klein would see these aspects of visualization as closer to a humanistic and interpretive approach, but Google Fusions also offers other graphs that do not embrace interpretation. There are bar graph and line graph options that I do not think Drucker and Klein would encourage the digital humanties to utilize.