Analysis
Overall, the Federalist Papers are an interesting collection of documents because of its density but also its reach to various topics of government and politics. These documents could be analyzed in numerous ways. For instance, network analysis would be a very useful tool to see connections to specific themes and topics. However, for this project, we decided to use MALLET and topic-modeling as our primary method to analyse these documents in order to help answer our research question.
Our Methods:
We chose approximately half of the papers chosen by Hamilton and chose them on the basis of content. Many of his papers were additions to a certain topic. So, to have the ability to use all topics, some were taken out for the sake of time. We identified thirteen element tags that specified the topic of content. These were features of government like the judiciary, executive, military, taxation, etc. Within those elements were four attributes: "elitist", "populist", "both", and "neither". These indicate the tone and type of statement/belief Hamilton possesses within a certain element. This combination is a way to track and analyze his views and opinions on certain aspects of government. We created our markup in XML, validated the papers via a RelaxNG schema, and used MALLET for topic-modeling, which would provide a visual of the content within the papers that was vital towards the comparison.
Elements
For this project, we assigned and identified thirteen relevant element tags that were the primary component of our markup. These are the following elements and their meaning:
executive: text tagged with this element related to any statement concerning the executive branch of the US government.
judicial: related to any statement concerning the judicial branch or court system.
legislature: any statement, set of statements, or phrases regarding the legislative branch or its implied powers.
military: any content relating to military action or its composition.
taxation: text tagged related to taxes, economic policy of the federal government, and opinions on historical views such as the excessive taxes that motivated a rebellious British colony.
trade: anything relating to commerce or economic activity that dealt with trade.
foreignPolicy: anything that related to diplomatic relations with other nations (which was mostly Great Britain), along with its varying philosophies and approaches to foreign policy.
liberty: any statement or phrase relating to individual freedom or rights. This not just applied to the general masses, but also the powerful few.
federalism: tagged phrases or statements regarding the powers of government and the separation of duties and powers of the federal and state levels of government.
historicalComparison: any content of the letters comparing the current situation of the colonies to Britian or other governments.
significantPhrase: writing that possessed key information that formed an opinion on a certain topic. These statements did not have anything to do with government, rather they implied a view relating to humanity or morals.
Attributes
We used four different attributes for our markup. They were:
Elitist:
Populist:
Both:
Neither:
Our Sources:
We used a variety of sources for the research and analysis of this project. We found an XML document of all 85 Federalist Papers from Github. Specifically, the federalist-bookworm: https://github.com/bmschmidt/federalist-bookworm. We narrowed our markup documents by utilizing the Library of Congress online resource/database of the Federalist Papers, which show the author and provide a small summary of each paper. From there, we marked up the 24 papers we chose with the elements and attributes created, along with changing the metadata and the root element. Next, we created an RelaxNG Schema and conducted a batch validation across all 24 papers.
We used governmental resources such as the Library of Congress's website that displays the corpus of documents to indicate which of the three founding father's wrote which paper. Because they all wrote under the pen name "Publius", latin meaning "the people", it would have been virtually impossible to know who wrote each paper. This was vital for our selection of documents for our analysis.