stenskjaer/dissertation-notebooks

Jupyter Notebooks containing analyses for my PhD dissertation

Doctrinal notebooks

This repository provides a set of notebooks used to investigate doctrines concerning intellectual self-knowledge in my PhD dissertation, Intellectual self-knowledge in Latin commentaries on Aristotle's De anima from 1250 to 1320: Qualitative and quantitative analyses. This repository provides much of the basis for the quantitative part of the subtitle.

The notebooks can be read in static form here, but there are also interactive versions of them available by clicking the badge "launch binder" above or following this link.

The notebooks are Jupyter notebooks, which means that regular running prose is interspersed with calculations in the Python programming language. This lets me think aloud and document what I do as the investigations go along. It is therefore possible to validate my process of thought and the details of any calculations, and in the interactive versions of the notebooks it is also possible to investigate the content of the database further. Any results presented in the main analyses of my dissertation from the use of the database should therefore be reproducible.

The data provided in these analyses is stored in a Neo4j database.

Database structure and content

A Neo4j database is a graph database, where nodes (also called vertices) are put into relation with each other by edges. Nodes and edges can be of different types and contains any number of key-value pairs for data structure and registration.

The database contains the following relevant node types:

• Doctrine: Singular points of doctrine.
• Instance: Single instance of a doctrine.
• Question: Individual questions in a particular question commentary (similar to a chapter).
• Text: Individual texts.
• Author: Individual authors of texts, including commentators.

The database contains the following relevant edge types:

• WROTE: An Author wrote a Text.
• CONTAINS: A Text contains a Question or a Question contains a Doctrine
• HAS: A Question has an Instance of a Doctrine.
• OF: An Instance is an instance of a Doctrine.
• ATTACKS: A Doctrine attacks another Doctrine.
• USES: A Doctrine uses another Doctrine.
• SUPPORTS: A Doctrine supports another Doctrine.

This results in the reduced database schema.

Properties on nodes:

• Doctrine:
  • uuid: ID of the doctrine.
  • description: The identifying headline of the doctrine. This is what is used as the title in the appendix of the dissertation.
  • note: A description of the node. This is used as the description in the dissertation.
• Instance:
  • uuid: ID of the instance.
  • type: Indication of the instance type. These are one of: ["Negative ratio", "Positive ratio", "Refutation"].
• Question:
  • uuid: ID of the question.
  • title: Question title in the commentary.
  • number: The question number in the commentary. This is no used often.
• Text
  • uuid: ID of the text.
  • title: The title of the text.
• Author:
  • uuid: ID of the author.
  • name: The author name.

These nodes make up the substance of the registrations. For some calculations it has been relevant to filter the material according to which chapter of my dissertation it is treated in, and I therefore also apply the node type DissertationPart. The database contains two examples of that node, referring to chapter two and three of my dissertation, with a relationship of type INCLUDES to each doctrine Instance.

Extra graph examples

Just to give an impression of how some elements of the graph look like, here are two examples that are also included in the introduction of the dissertation.

Example graph of the presence of a doctrine in six different texts.

Example graph of arguments given by John Dinsdale on the science of the soul.