Final Portfolio Revision and Reflection

One of the most important lessons I learned this semester was the vital importance of protecting the integrity and originality of historical evidence. It wasn’t that I had not valued this aspect of historical research before, but this class opened my eyes to the hidden dangers and avenues in which data manipulation can have a significant impact on the way we perceive research. I am now much more confident not only in my abilities to recognize these errors in data collection, but to utilize the tools available to me to prevent such manipulation. When I say manipulation, I do not mean data handling with a malicious intent (although the existence of such “research” methods certainly do exist). Instead, I have learned the value of recognizing the process in which we, as digital scholars, must handle historical data – with as much care as we would a real-world artifact. Data is not as sound as it may appear, its fragility (especially in sets that measure beyond our capacity for close analysis) is as corruptible in the wrong hands as it is in the right ones.

This leads us, as researchers, to process the way in which we analyze such data sets, and make note of our influence on the final product. One of my favorite terms I learned this semester was “derived data” because its implementation into the fields of history and digital humanities may be understated in how we come across “new” information. As a humanities major, perception and interpretation have always been a particular focus of mine and when entering the quantitative realm of digital scholarship, it seemed that these methods would no longer be necessary when studying simplified analytics. While I learned this could not be farther from the truth, I also found it fascinating that as much effort should be put into studying our processes of historical and digital research as into the project itself. I believe there is no better way to prove this fact than the construction and culmination of this portfolio. I have always bore a great disconnect with academic literature and it’s lethargic pacing in regards to proving an argument. While academic literature is still boring, I do have a greater appreciation for the effort scholars make in detailing their process of historical research. I suppose I would ask myself when constructing an academic argument how future researchers would analyze my work and if I would do them any favors by explaining my own thought process.

This brings me back to the concept of “derived data” and how it affects our perception of historical research. We have literally thousands of years of historical research that has been collected and is now being transposed into the digital realm. As historical scholars it is first our main priority to preserve these findings for their future use. That future use is our second priority in how we present these findings for the rest of the world. We looked at many websites, archives, and digital media this semester – all of which was presented in a manner that best suited the purpose of the author. It is a not so subtle realization that the information we absorb is from another person’s ideology of how that information should be transferred to us. Data points are placed on a chart for us to visualize and better comprehend, but the simple fact that chart exists is proof we have bought into a molded idea of how that data should be understood. Derived data encompasses not only logical assumptions we can make about a particular data set (this would be metadata, specifically) but its presentation and subsequent absorption into the minds of other researchers.

I thought about this process of historical research quite a bit when revising my visualization project. The activity I chose to revise was my Week 14 data set detailing the Washington’s Enslaved Peoples in 1799. Originally when I did this project, I just created a data set for the enslaved peoples in the Mansion House. and only looked at name, age, and owner. My revised edition now includes records of enslaved peoples in Muddy Hole, River Farm, Dogue Run, and the Union Farm. I also added the profession of those that were specialized for a particular trade and their current work status. My original visualization in Flourish did not turn out as I had planned and I had to make adjustments within the data set to fit the visualization software. Recalling this process reminded me of data manipulation for the accommodation of visualization. It is the responsibility of the researcher to respect the integrity of the data and find ways to present it in its truest form, not to whittle away and mold it to fit into our own form of comprehension. My newest visualization attempts to better support the data set and shift its focus from accommodation to proper reflection.

[Revised] George Washington Enslaved Peoples 1799 copy

I made these visualizations in Tableau. In the first visualization, I show how many enslaved peoples are at each of the Washingtons’ estates and farms, and how many of those are owned by George versus Martha. I believe this visualization compares nicely how each farm is supported by each owner, while also telling just how many slaves the Washingtons’ owned. In the second visualization I show the work status of all the Washingtons’ enslaved people. The highlighted “Null” column shows all those without given comments about their status. It was surprising to me how many children were owned by the Washingtons (easily 1/3 of all the Washingtons’ slaves were children). The final visualization shows how many of the Washington’s slaves were specialized in a particular trade. It would be interesting to see this data set after about 5 years and see how much would have changed.

As I close out my senior year and look to the future, I hope to continue to work in the digital humanities either as a digital scholar or specialist in data analytics. I believe the tools I gained in this class and others have made me more aware of the troubles that may lie in data research that are so often and easily overlooked. The future of digital humanities is bright and I hope to join in the discussion as our technological world moves ever forward. I am also excited by the prospect digital means of research can mean to fields looking to expand their research horizons. History has always been a favorite subject of mind and to be given the chance to explore this field through the computational lens was a fascinating experience. I can see nothing but good things ahead when I think of the skills I have learned in relation to digital, historical, and analytical ways of thinking.

14. April 25 – Data Visualization

George Washington’s List of Enslaved People’s 1799

My data set: George Washington Enslaved Peoples 1799

My visualization:

I used the transcript of George Washington’s Enslaved People’s in 1799 as my data set for my visualization. I used the records from the Mansion House and identified name, age, spouse, and ownership. After inputting my data set into Flourish I immediately noticed a slew of problems. It was reading each entry numerically which meant I had to change spouse entries from yes or no to 90 and 0 to accommodate. I did this for both George and Martha as well. I think my revision will look to adjust this accordingly or perhaps use a different software.

13. April 18 – Data Visualization Critique

At the beginning of the semester, I used to rely solely on data visualizations and textual analysis to receive information and study research. However, through the DH curriculum, it has become more apparent that data integrity is best protected by limiting our reliance on digital tools. Just as qualitative data can be misinterpreted or construed, quantitative is just as easily fickle when attempting to best translate its substance to a more comprehensible form. Charts and Graphs in particular play effectively with our preconceived notions of importance and relativity to better our understanding, but can also be used to manipulate or confuse how we should think about these data sets. Color, weight, size, movement, and shape can all be tools used to display data in a certain fashion that may or may not reflect the comparative nature being presented within the data itself. Also, pie graphs are bad.

12. April 11 – Data Analysis

When it comes to analyzing anything, the more organized and structured the subject is, the easier it may be to conceptualize information about the subject, and limit certain self-perceptions and biases. Tidy data assists historians in this way by narrowing their focus on core points of interest pertinent to a broader narrative, event, and/or system. All this, however, weighs heavily on the structural integrity of the primary source, aka received data. If the received data lacks its own structure and organization, the modern bias of the historian is imposed (to an extent inadvertently) to craft the data into a more understandable format. It is much more truthfully reflective of the primary source to contain its own categorized elements representative of the subject and time-frame of the subject. When faced with information holes, however, analysts create derived data – data created in corroboration with other outside points of information to fill these holes. When asking historical questions, it is imperative to discern what is received and what is derived. The unique perceptual interaction with data from one analyst to the next can skew and manipulate a primary source beyond its raw, intended values. It then becomes the responsibility of the researcher to infer what they can from received data in the subject’s best interest – meaning, how can we as historical researcher’s protect the historical integrity of a primary data source while expounding our own modern, narrative structure in a way that both projects the historical evidence in a truthful light while also revealing additional information for future analyzers? Our encounter with contemporary data, derived data, and metadata must then also reflect this goal, as our responsibility to the primary source can be validated through the work of secondary research. Only then can we continue in our own research and analysis to bring about our own unique perceptions and findings.

11. April 4 – Data Map

For my map, I decided to look at the surrounding counties of the early universities in the United States. I was curious to see how prevalent slavery was in these areas as most colonial universities in the US were built by slaves. I looked at the 9 earliest universities of the US: Harvard, William & Mary, Yale, Rutgers, Columbia, Brown, Penn, Princeton, and Dartmouth. These universities were all completed and operating by 1790, so that was the data set i chose to utilize. I had to research the counties which these universities are located in, and every one except Penn’s was a legally recognized county at that time. Rutgers extends to both York and James City county, so I included both.
My visualization revealed some interesting findings. Harvard’s county, Middlesex, was recorded as having 0 slaves by 1790 (so this may require additional research to prove). There is also a very apparent increase in slave numbers the further south you move down the map. For instance, enslaved peoples recorded numbers increase by over 1000 once you cross the Mason-Dixon line.
What this visualization doesn’t show us, however, is how these counties compare with those nearby. It would have been better to find a way to highlight these counties without removing the others. It also would have been good to see the numbers of enslaved peoples at the time of each universities construction rather than all of them in 1790.

10. March 28 – Data Map Critique

Data-driven visualizations should be considered an interpretation of the data set they emulate rather than a direct and 100% accurate reflection. As digital scholars we look to find ways to comprehend these complex sets of data and are able to do so by visualizing them on a  map. However, the simple fact that we use an outside tool, such as a map, is indicative of our own implementation to understand what has already been given. It is through these methods that we manipulate data to better construct the narrative we wish to tell. It is crucial for researchers to recognize and acknowledge this process as it is the duty of the data analyst to protect the integrity of the data set. It is just as easy to manipulate data and lie with a map as it is to go great lengths to prevent such incursion by implementing alien methodologies with the visualization process. They are not of equal moral repercussions but equally distort the data to fit our own comprehension.

5. February 14 – Narrative and Scholarly Communication

“Lest We Forget: The Triumph Over Slavery”
– Digital Narrative
– Content: The scholarship is sound but uses very common phraseology and does not include citations or quotes within its text. It is well communicated for most people to understand, although most summaries are brief and do not have too much specific information.
– Design: The layout of the site is very straightforward with a timeline of events at the bottom for easy navigation. There is great use of images and graphics. Although the site is generally easy to navigate, it requires FlashPlayer to operate and most likely would not be available on a mobile device. The type is small and the font is faint and boring. When it works, however, this project is fascinating to navigate.
– Audience: This project is targeted towards a non-academic people looking to gain brief knowledge about the history of slavery. Its accessibility and simple language help it achieve this.
– Digital Media: This site is run completely by digital media, so without the proper software to run it, one could not even access it. That being said, it utilizes this format well, containing image movement and transitions. It does not, however, contain any video or audio files.
– Creators: This projected was curated by the New York Public Library, particularly the Schomburg Center for Black Culture. There is a credits page listing contributing memebers and researchers.

“Frederick Douglass National Historic Site, National Park Service”
– Archive
– This site contains a plethora of primary sources and objects. The archival aspect of this site is well-communicated to users as a simple digital database of an actual museum.
– Design: This is a Google exhibit template but uses its simplicity to be as accessible an open as it needs to be. It clearly lists exhibit items and catalogue in a non-confusing way. Navigation is simple yet not necessarily specific. Everything operates as it should although there could be more specifity in identifying certain items. This website does not need FlashPlayer and works well on a mobile device.
– Audience: There is not a clear audience being targeted, and the site’s layout does not lend itself towards a particular narrative structure. This site would be best suited for those with a specific interest in this museum, or scholars looking to research a particular item.
– Digital Media: This site does an excellent job of displaying its catalogue through means of digital images. It integrates a map feature to help users find the actual museum. It goes great lengths to digitize the real world and cross reference both entities, even without the use of video and audio.
– Creators: This site simply tags the National Park Service as its contributor without mentioning any individual people or other institutions (at least it does not in any immediate or accessible link).

4. February 7 – Annotations

Annotating these sources was particularly difficult since I chose very specific sources to deal with. One thing I found interesting was researching the growth, expansion, and evolution of the manufacturing company to its present day company. My classmates had great sources and used annotations much more effectively than I did.

2. January 24 – Research Questions

I’m interested in how the Civil War affected the process of shipping letters to and from Frederick Douglas. To research this question I will need to find more primary sources on the movement of mail throughout the US during this time period, as well as Douglas’ movements. Did he change addresses? How did the war affect postage in general? Were there a separate post systems in the confederacy or even in the union? Once I narrow a specific year for Douglas’ whereabouts, I can then focus on the movement of mail to and from that location. I will have to use this specific each time he moves and should note that in my research question. I have enough data about the dates the letters, and who they were sent from, however it will be difficult to analyze their movement without a form of digital map showing me where they came from. I will have to expand my research into identifying maps of that time period and use them to help further my question.

1. January 17th

John Brown

– Insurrection is the buzzword of the mid-1800s

– 2 sources from a Southern-Democrat point of view and 2 from a Union-Republican.  Both illustrate the arguments and biases of th two parties at the time.  The Democrats blame the Republicans for instigating John Brown’s actions through their radicalization of emancipation.  The Republicans do not, in turn, defend John Brown, but instead deflect his association with the party and blame the Democrats for bringing the Harper’s Ferry raid upon themselves.

– The excerpt from Lincoln puts it best in which he says, “Your purpose, then, plainly stated, is that you will destroy the Government, unless you be allowed to construe and enforce the Constitution as you please, on all points in dispute between you and us.”  This political discourse, I think, best illustrates the moments building to the Civil War.

– Conversely, this excerpt from a Southern newspaper instigates the “Northern Aggression” perpetuated throughout the South leading up to the Civil War: “It fully establishes the fact that there are at the North men ready to engage in adventures upon the peace and security of the southern people, however heinously and recklessly, and capable of planning and keeping secret their infernal designs.” – Charleston Mercury

  • In 1859, this was written in response to John Brown’s Raid: “It is the presage of the future storm, that shall desolate the whole land, if the people give this abolition doctrine their approval. It necessarily tends to servile insurrection, civil war and disunion.” – Chicago Press and Tribune