ABOUT
TECHNICAL
The three technical levels of a Digital Humanities project are outlined below, accompanied with the steps we took along the way in order to accomplish this project.
SOURCES
Sources are raw materials, including files, images, texts, and sounds, used in the course of research (Turabian, 2018). Completely original digital humanities projects result when researchers commence a mission to uncover significant data about a certain subject by engaging with sources, whether that be paintings at the Tate Museum or public prisoner records from the Eastern State Penitentiary.
For this project, we did not have to compose an original data set. We decided on the Eastern State Penitentiary records from the list of 25 available data sets given by Dr. Garcia. We chose it because we were intrigued as to the possible questions we could ask about the inmates, and how information from almost two hundred years ago could relate to today's age of mass incarceration in the U.S., where nearly 1% of the population sits behind bars and ethnic discrimination remains an issue in criminal justice (Prison Policy Initiative, 2016).
Data is a piece of information, often collected together for a specific purpose. It can be quantitative or qualitative, empirical or theoretical. After sources are compiled and reviewed, workable data results. A data set is constructed through given data that's compiled based on trends or some other criteria for collection. Quantitative or empirical data is objective, while qualitative or theoretical data is subjective. The ESP records consist of both quantitative and qualitative data. Dates of admission, sentence lengths, and literacy levels are examples of quantitative data, which are represented with numerical values. Qualitative data is found in the columns covering Ethnicity, Religion, Occupation, which are categorical data, or that which can be broken into groups.
In order to provide context to the subject of the project, researchers must conduct an extensive literature review by mulling over extraneous literary resources somehow related to the topic of the data set. Our team employed books and journal articles in order to interpret, analyze and build ideas about our research topic. They can all be found in the Annotated Bibliography section of the project website. The books provide a more detailed and illustrative representation of the Eastern State Penitentiary. The journal and newspaper articles narrow down specific parts of our research questions and were used to address certain nuances. Most of our sources were available in digital format, with the exception of a few hard copies of books checked out from UCLA libraries. Through the literature review, we obtained a more comprehensive understanding of the architecture, personnel, and administration style of Eastern State. After learning that Quaker religious views guided the funding for and construction of the penitentiary, we became curious to know how religion influenced its operations (Johnston, 1994).
The research problem guides the search for evidence, models, and arguments to respond to (Turabian, 2018). Our research questions, including how religion influenced the reform methods of Eastern State inmates, and how the then-revolutionary design of solitary confinement affected the prisoners, led us to look for evidence, particularly through text analysis, of religious terminology and themes. The research problem points us to the next level of processing.
PROCESS
How do you make your source analyzable?
This means getting the sources into the computer and making it computationally tractable. The course in doing so consists of setting your agenda and choosing testable hypotheses. We set an agenda to reveal the connection between religion and inmate rehabilitation, which were guiding principles in the construction and formation of Eastern State.
Data cleaning is an arduous task involving tasking through the categories of the data in order to allow digital tools to interpret it more easily and clearly. We cleaned the data by separating the category of ‘RaceEthnicityOccupation’ into three categories to better analyze some of the defining characteristics of the inmates. We also quantified their individual literacy levels by coding them as 0 or 1, illiterate or literate. We distinguished the crimes that they committed into violent and non-violent, in order to see the overall trends of criminal activity that would have landed one in Eastern State, and additionally to judge whether solitary confinement was really a just punishment for non-violent offenses. Finally, in cleaning the dataset, we created a new category, defining the inmates’ level of hopefulness of redemption according to the moral instructor's descriptions.
Metadata is data about data. It can point to more interesting and idiosyncratic trends within a larger dataset (Riley, 2017). For example, when coming up with ideas of how to interpret Eastern State data, we found trends in the moral instructor's descriptions that led us to identifying his virtual prognoses for their redemption based on Christian ideals.
Data silence is the absence of information in a set, which is important because it tells us about how the data was collected, who collected it, potential inherent biases, and it can point to the attitudes of the time (Padilla, 2018). It tells us the record of history is always shaped by personal narratives and interpretations as well as verifiable facts. In our case, a large number of inmate records were missing information about ethnicity. This is evident in the visualization on the "Prisoners" page of the website, categorizing hopefulness and hopelessness by ethnicity. This affects our overall understanding of how we perceive Eastern State's treatment of prisoners based on their race or ethnicity.
There are a multitude of tools available for processing data. Data visualizations are useful methods of displaying researchers' findings in a comprehensible way. There are different visualizations that are more beneficial to certain types of data. Acceptable charts for categorical data include bar graphs, pie charts, symbol plots, tree maps, and mosaic plots. We used a tree map to plot out the proportion of different crimes committed by the inmates. Line charts are used to express probabilities, or compare two or more values, and work well with time-based data. We employed this type of graph in the timeline tracking admission rates throughout the decade. After playing with all of the digital tools exposed to us in class and lab section to analyze different components of the data, we narrowed in on the more telling few to include in the project. The visualizations we included were made using Palladio, Tableau, Voyant Tools, Raw Graphs, Google Sheets, and ArchGIS. After reviewing the literature, cleaning the data, and creating data visualizations, all of that information needs to come together in a pleasing way, which brings us to the final level of a DH project.
PRESENTATION
How do you present all the work that you've done to your sources?
The design of a site or other digitized work becomes the 'face' that the viewer or listener encounters to comprehend your data and research. In choosing a presentation style, it's important to identify your audience. The target audience for our project is the general population interested in the history of the 19th century, particularly scholars and students interested in exploring how religion influenced the state’s reformation of prisoners, and the revolutionary system of the Eastern State Penitentiary. It may be helpful for people researching incarceration trends in America in regards to race, gender, and religion. This project may answer questions about how the Pennsylvania system has influenced contemporary models of incarceration, particularly systems that utilize solitary confinement and that use religion as the basis of their reform program.
In preparing an authentic presentation that best portrays the ideas and interpretations we decided on, its important to keep in mind the themes that can define the project. The theme of power is prevalent in all digital humanities projects. It's the ability to influence or effect change. Those with power can significantly influence how data, or information, is collected and interpreted. Prison administrators and personnel, including the moral instructor, had the heaviest hand in molding lasting impressions of these inmates as we've observed through their written records.
The narrative is the story that becomes realized as data is interpreted by those in power. The narrative defines how we perceive the record of facts and evidence. However, the story is never entirely complete because information may be missing, or silent. Overall, power shapes the narrative which is constructed through both data and silence. History is an imperfect record of events throughout time. The records kept at Eastern State Penitentiary represent both public and personal narratives. We've taken official state records documenting the public history of Pennsylvania's incarceration, and interpreted personal notes to build the narrative surrounding inmates' perceived chances of redemption through analyzing their empirical classifications. This process leads us to an amalgamation of actual events or instances that happened as well as the narratives we have about them.
Thus, the production of history is effected by inclusion and absence, which builds an imperfect record. In this sense, our project on ESP and conclusions we've drawn adhere more closely to the constructivist theory of knowledge that says knowledge consists of both the facts and the narratives about those facts. Our understanding of Eastern State Penitentiary was shaped by a narrative which we've constructed via thorough data analysis, and we are happy to share that through a presentation in the form of this website.
TEAM + ROLES
Leandro Assuncao
Leandro Assuncao is a fourth year Anthropology major with focus in Socio-Cultural Anthropology. Leandro used Norman Johnston’s Crucible of Good Intentions, Noel Ignatiev’s How the Irish Became White, and Charles Dickens’ American Notes as reference to his texts. The spider map that references the prisoners’ birthplace to their sentencing locations was produced by Tableau. The lines connecting birthplace to sentencing location was served to show in a larger dimension that each prisoner went through an individual journey to get to the same place behind bars. Leandro also created a side-by-side bar graph to compare sentencing years. A side-by-side graph helps to compare information from one category to another.
Kelsey Boyle
Kelsey Boyle is a fourth year Communications major and Digital Humanities minor. She helped clean the data to categorize prisoners as being “hopeful,” “hopeless,” or “demented.” Voyant Tools was used to do a text analysis for this qualitative data, which helps convey the effectiveness, or lack thereof, of Eastern State Penitentiary. She worked on the text analysis web page as well as helped format and design the website.
Jiarui He
Jiarui He is a fourth-year Economic major and Film minor. She annotated John H Kramer and Jeffery T Ulmer’s Sentencing Guidelines : Lessons From Pennsylvania, which is on the annotated bibliography page. She used Rawgraph to create a treemap displaying offense type of 519 prisoners. The space in the visualization is split up into rectangles that are sized by the number of prisoners for each convicted crime. She also wrote a corresponding paragraph to give a written description of how prisoners are distributed into different crimes.
Frances Keady
Frances Keady is a fourth year Political Science major and Digital Humanities minor. She annotated the articles found in The Prison Journal and Pennsylvania Journal of Prison Discipline and Philanthropy, providing insight to the institution's religious history and common labor practices in the state prison system. She contributed to cleaning the data by separating the categorical sections covering race, ethnicity, occupation, and levels of hope. She created initial drafts of the project website and wrote the the "About" and "Today" pages on the final website.
Ly Nguyen
​
Ly Nguyen is a fourth-year International Development Studies major and Entrepreneurship minor. She used Tableau to create the timeline for Admission Rate by Ethnicity overtime and ArcGis to create the heat map of sentencing locations, to emphasize a large number of prisoners coming from Philadelphia. She also referenced Johnston’s Eastern State Penitentiary and Wintermute’s Crime and Punishment in Eastern Pennsylvania to provide context to her visualizations.
Amanda Norman
Amanda Norman is a third-year Communications major and Digital Humanities minor. She annotated Muriel Shmid’s The Eye of God and Thomas and Zaitzow’s Conning or Conversion? The Role of Religion in Prison Coping, both of which can be found in the Annotated Bibliography section on the website. She also created the visualization found in the Prisoners section entitled, Hopeful vs. Hopeless by Christian Impressions. This bar graph shows the number of demented, hopeful, and hopeless prisoners who either had Christian impressions, no Christian impressions, or no mention of religion in their descriptions.
ACKNOWLEDGEMENTS
We would like to thank our professor Dr. Garcia and our TA Dustin O'Hara for exposing us all to the field of Digital Humanities and showing us how to use all of the tools we've incorporated into this project. We would also like to acknowledge the team that works to maintain the Eastern State Penitentiary Historic Site in Philadelphia and the scholars that update the Penitentiary's Research Directory, without which we would not have been pointed to our most crucial sources in understanding the history and impact of Eastern State Penitentiary.