• Replicating Fortier's THEME System for Digital Text Analysis

    Author(s):
    Kaylin Land (see profile) , Geoffrey Rockwell, Stéfan Sinclair
    Date:
    2020
    Group(s):
    CSDH-SCHN 2020
    Subject(s):
    Digital humanities, Symbolism, France, Text data mining
    Item Type:
    Conference paper
    Conf. Title:
    CSDH/SCHN Digital Humanities Conference 2020
    Conf. Org.:
    CSDH
    Conf. Loc.:
    Online
    Conf. Date:
    June 1-5, 2020
    Tag(s):
    Coding, Historical replication, Python, French symbolism, Text analytics
    Permanent URL:
    http://dx.doi.org/10.17613/k1fn-n824
    Abstract:
    In 1971, Paul Fortier created a computer program to save significant time in analyzing French literary theme words connected to semantic fields. The system, aptly called THEME, harnessed the capabilities of computer-generated keyword concordances with frequency and distribution calculations to create research reports for user-defined literary themes. Fortier's system represents a significant achievement for digital humanities, not only due to its impressive capabilities but also for the precedents the system created in conceptualizing the role of the computer in text analysis. This paper discusses efforts to recover the THEME system and create a working approximation of the system in Python. This effort is part of Stéfan Sinclair and Geoffrey Rockwell's Epistemologica project that seeks to recover, valorize, and interpret historical text analysis in the humanities.
    Notes:
    Please use this link to interact with our paper via Sway: https://sway.office.com/GQhUyvFkofaoNu6h?ref=Link The file associated with this entry is a .pdf version of the Sway page. View it online for increased interactivity and to view a short video presentation.
    Metadata:
    Status:
    Published
    Last Updated:
    3 years ago
    License:
    All Rights Reserved

    Downloads

    Item Name: pdf replicating-fortier′s-theme-system-for-digital-text-analysis.pdf
      Download View in browser
    Activity: Downloads: 78