ABOUT

TEAM MEMBERS

DR. WILLIAM CAMPBELL - INSTRUCTOR

Digital Humanities Instructor.  

ANDREW WESTON - QUALITY CONTROL OFFICER

Information Technology major.  

AARON ZAVATCHAN - PROJECT COORDINATOR

Anthropology and History major.  

COLE BALL - SUBJECT MATTER EXPERT

History major.  

GIA SALVIO - DOCUMENTATION SPECIALIST

History major.  

WILLIAM DELLER - WEB DESIGN/USER EXPERIENCE

Information Technology major.  

XML Markup

The raw data for our project came from the speeches delivered by each of the 3 main categories of Civil Rights activists: nonviolent, the NOI, and the BPP. These speeches either already existed in an online transcript, or the group member marking up that speech had to listen to the speech’s recording and make a rough transcript of the speech itself. But for the most part, the speeches we used are well known and therefore have already been typed up online.

The next step from getting the speeches themselves and the copyright information in an xml document was then to frame the speech with some basic tags. These were identifier tags that included the speech title, the speaker’s name, the speaker’s organization (NVM, NOI, or BPP), the speech’s date, and then the source from where the speech was found. These xml markups are stored in our XML folder on our repo. Using regex for most of the speeches, the paragraphs were tagged using a close-open strategy.

The XML markup posed a challenge because a lot of the things that were tagged in the speeches are subjective. Out of the 5 group members, I think all of us tagged at least one part of a speech, so there were different perspectives when tagging the speeches. Some of the nonnegotiable tags were of course the p tags, then the people tags which was a simple tag that gave the name of the person in an attribute value.

The more complicated tags include set, org, cause, and acts. The set tags defined each of the groups listed in a speech. This categorization could include anything from an ethnic or racial group, a religious group, and age-defied group, and an occupational group. The cause tags defined a category of which the speaker was referring to. This could tag class unity, segregation, integration, schooling, and more. This is however, different from the acts tags because the acts tags identify direct actions on the behalf of each speaker. The acts are defined by the groups of Civil Rights activism our group set out to compare from the start: Nonviolence, active resistance, and self-defense. The org tags identify an organization that each speaker refers to. This includes democratic groups such as Democrats and Republicans, organized groups like SNCC, CORE, and the NAACP, and religious organizations such as churches. The org tags are meant to be used when referring to more organized groups.

One of the newer tags that was added included opinion tags for the orgs mentioned above. The purpose of an opinion tag was to take that data and create a visual graphic for the website that would demonstrate the relationships between the speaker mentioning an org, and the org itself. The opinion tags were rated on a scale of 1-5 which again, was subjective to each of the people tagging the opinions. Starting from 0 which meant a terrible opinion and therefore a negative relationship, 3 being neutral, then 5 being a great opinion of the org therefore a great relationship.

Visual Representations: Network Graphs, Histograms, and a Timeline

Network Graphs

This work was done by William and Aaron who spent time going through each of the speech tags and first figuring out a way to represent the things they wanted to. Their final decision came with Aaron identifying the opinions with a numerical system in the organization tags in each of the speeches, then William working on the graphs themselves in Kumu. They both worked out a color system as well which indicates relationship as well. The arrows also show relationship by proportion. The larger the arrow pointing from the speaker to an organization, the more the speaker mentioned the org.

The network graphs our project utilized are mostly to represent the relationships between each of the speakers vs. the organizations they speak out. This ties into the opinion tags I talked about earlier which were used to directly influence the graphs themselves. So, since the data about each speaker comes from the speeches themselves, that is where the data from the graphs comes from. There is no external outsourcing of information. Whatever information was given in the speeches is how the graphs were created. Each speaker has at least one speech, and under the tab where the group each speaker was in is listed, is the speaker and then their subsequent speeches that they gave that our group analyzed.

Each speech has its own network graph as well, even if given by the same speaker. There is a more in-depth explanation for each graph in the pages themselves under the “explanation” section.

Histograms

The histogram was worked on by Cole and Andrew, who aimed to, as the title of the histogram says, visually represent the “Frequency of Acts by Organizations”. The two decided at first to make a graph that included the acts that were tagged as nonviolent resistance, self-defense, or active resistance from each of the 3 groups: Nonviolence movement, the NOI, and the BPP. This was done by creating an xquery file which was turned into a sag file that gives the final results as seen by the histograms.

The active resistance was shown as a red portion of the graph, the self-defense was yellow, and the nonviolence was green. Each of the bars adds to 100%, and the histogram allows you to see how much each group talked about each act.

There were changes made from the first draft of the histogram. The first histogram included 3 bars which grouped all the speakers under each group to form the NVM, NOI, and the BPP. After some discussion, Cole and Andrew decided to go more in depth by showing how much each individual speaker from each group references the acts. This way you could see even the differences between the group member’s themselves, such as Malcolm X vs. Ali, Dr. MLK vs. Jackson, and Newton vs. Hampton. Just as with the above network graphs, the histograms are going to be put on each page of the website under the respective groups pages, and more information under the about graph section will be there as well.

Timeline

The final visual representation our group included in the project was the timeline, which Gia worked on with help from William and Aaron with formatting.

The timeline aimed to encompass the major years of the Civil Rights movement and include many different things that were relevant to our project. In the earlier stages, I had about 10+ pages of notes for the timeline, which I ended up cutting most out of for the final product. The final draft of the timeline spans from 1954 with the landmark Brown V. Board decision, to 1974 with the resignation of Richard Nixon. William formatted the css of the timeline html page to frame it in the same formatting as the rest of the website. Aaron added the tooltips where if you hover over the timeline event, a box pops up on the right which gives a description of the event.

The major categories I included on the timeline were the speech dates from the speakers, Supreme Court cases, major legislation, assassinations, the inauguration dates for some presidents, and Vietnam War draft information such as the beginning date of the draft, when the final draft was held, and the day Muhammad Ali refused the draft.

I tried to keep the formatting the same across all of the timeline by using past tense for the events such as “…case decided” or “president inaugurated”, or “…assassinated”, etc. Also to keep formatting even across the board, I chose when a Supreme Court case was decided, and when the legislation I mentioned was ratified.

We at first tried to keep the timeline from 1960-1970 but I realized there was too much information happening before 1960 and shortly after 1970 to be cut out, hence why the timeline extends past the decade. I didn’t include images but that was a preference because I wanted the audience to focus on the content itself rather than be bogged down with oversaturated content like a hundred images. This is also sort of the reasoning behind not flipping the timeline to go left and right, I thought a middle anchored timeline helped the reader just go from top to bottom which I thought was easier.