Skip to content

lokalkosmos/Lesprit

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

52 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

The Digital Dialectica and Montesquieu's Book XIX

Website: https://the-digital-dialectica.rarebook-ubfc.fr/
Licence: MIT

A critical Digital Humanities project examining Montesquieu's De l'esprit des lois Book XIX through computational methods that expose the incompatibility between contextual humanistic philosophy and digital infrastructure that is built for purposes other than humanistic understanding.

Author: lokalkosmos
Institution: Université de Franche-Comté


Table of Contents


Project Overview

This repository supports a Master's thesis examining Montesquieu's De l'esprit des lois Book XIX, specifically Chapters I-VI where he articulates the concept of esprit général (general spirit). The project employs Digital Humanities methods as objects worthy of critical analysis themselves, investigating what happens when computational approaches and ideologies are applied to the humanistic learning of philosophical concepts that in principle contradict the ideological frameworks and operations such methods require.

The central argument: When digital tools, computational logic, and Big Tech ideologies, rhetoric and practices meet Montesquieu's philosophy of irreducible contextual complexity, the resulting friction generates insights about both eighteenth-century political thought and contemporary digital scholarship. The project examines how computational methods originally built for military command, control, and corporate optimisation deal with philosophical commitments to humanistic contextual particularity that challenge systematisation.

Core Philosophical Position

Digital tools offer useful capabilities but remain fundamentally limited instruments requiring critical humanistic interpretation. This project treats computational limitations as philosophically significant findings within the ideology and rhetoric that supports an ever digitalising world. The tools examined here emerged from military and corporate infrastructure designed for extraction and standardisation. When these tools prove to be inadequate for capturing Montesquieu's contextual epistemology, it is a sign of their suitability for their intended purposes other than nuanced humanistic scholarship.


Research Philosophy

This project adopts a critically aware approach to Digital Humanities:

✓ Treats digital tools as objects of critical analysis alongside historical texts
✓ Acknowledges that contemporary digital infrastructure emerged from specific origins that shape its operations
✓ Recognises that abstraction shapes what becomes visible and what remains obscured
✓ Validates computational outputs through sustained human interpretation grounded in disciplinary expertise
✓ Embraces the friction when methods deal with texts that challenge their ideological assumptions
✓ Documents methodological struggles as research findings
✓ Examines how forced machine-mediation reduces the lived experience of contextual difference that Montesquieu's esprit général concept requires

The project rejects:

✗ Claims that computational methods are objective or neutral
✗ Technological determinism or solutionism
✗ Reducing humanistic enquiry to pattern recognition
✗ Black box analysis without acknowledging opacity and its implications for the ideology supporting an ever digitalising world
✗ Treating scale as inherently superior to interpretive depth
✗ Digital methods as inevitable rather than ideologically shaped tools

Following scholars like Johanna Drucker, Alan Liu, Stephen Ramsay, and Paul Edwards, this project understands that all data visualisation is rhetorical, that tools encode political possibilities through their design, that infrastructure carries forward the logics of its origins, and that Digital Humanities requires cultural criticism of its own practices.


Repository Contents

The repository contains transcription files, processing notebooks, datasets, and analytical outputs. All files are available at https://github.com/lokalkosmos/Lesprit



Corpus and Data

Primary Textual Corpus

15 editions of Book XIX (Chapters I-VI) transcribed and analysed:

  • Languages: French (original), English, Italian, German, Polish
  • Date Range: 1748 (Geneva first edition) to 1803
  • Method: Transkribus OCR with extensive manual correction
  • Format: Structured text files and JSON with hierarchical organisation

These editions enable examination of:

  • Textual variation across printings
  • Translation practices and semantic shifts
  • How esprit général changes across linguistic boundaries
  • Editorial interventions in different publication contexts

Extended Corpus

~100 editions mapped geographically:

  • Coverage: 17 European cities across 11 regions
  • Period: 1748-c. 1800
  • Source: Penny University dataset and institutional archives
  • Purpose: Demonstrating dissemination patterns whilst acknowledging that geographical coordinates cannot capture cultural complexity

Timeline Data

30 historical events (1689-c. 1800):

  • Montesquieu's biography and publication history
  • American and French revolutionary applications
  • Catholic Church censorship (Index, 1751)
  • Constitutional developments influenced by De l'esprit des lois

Methodological Approach

Transcription Process

Challenge: Eighteenth-century typography poses unique challenges for modern OCR systems across languages.

  • Tool: Transkribus for initial automated transcription

  • Problems Encountered:

    • Long s (ſ) misrecognised as f, s, or unrecognised
    • Ligatures (æ, œ) unreliably processed
    • Orthography and punctuation marks misinterpreted
  • Solution: Extensive manual page-by-page correction comparing OCR output against original scans.

This labour-intensive process shows a fundamental reality. Computational efficiency has limits in automation and cannot achieve scholarly accuracy without sustained human intervention grounded in disciplinary expertise.

Data Structuring as Epistemology

Converting historical texts to JSON and structured formats involves interpretive choices with implications:

  • Datafying nuanced and complex contexts and ideas into discrete computational units forces them into predetermined structures which may not accurately represent
  • Material features that add context (typography, layout, format, unique marks, binding) become invisible in computational textual analysis

Data preparation itself enacts the abstraction Montesquieu's philosophy problematises. The requirement for standardisation and uniformity contradicts his insistence on contextual particularity.

Large Language Model Analysis

  • Tool: Claude 3.5 Sonnet (Anthropic)
  • Purpose: Translation and semantic comparison across 15 editions

Critical Acknowledgement:

✓ LLM analysis operates as a black box that is fundamentally opaque to external examination
✓ Outputs probably contain subtle inaccuracies, anachronisms, biases, and generalisations
✓ Statistical pattern recognition optimised for engagement metrics does not equal philosophical interpretation
✓ Systems designed for surveillance capitalism's extraction purposes cannot accommodate humanistic nuance by design
✓ All automated analyses require validation through traditional scholarly expertise

Why use LLMs despite limitations? To simply test them. The opacity and inadequacy are documented as methodological evidence demonstrating epistemological limits and the ideological purposes these systems serve.


Technologies Used

Core Tools

  • Transkribus - OCR for historical documents (READ-COOP consortium)
  • Claude 3.5 Sonnet - Large language model (Anthropic)
  • Python 3.x - Data processing and analysis
  • Jupyter Notebooks - Reproducible analytical workflows

Libraries and Frameworks

  • Pandas - Data manipulation
  • JSON - Structured data storage

Visualisation and Web

  • WordPress - Content management (Université de Franche-Comté hosting)
  • TimelineJS - Interactive timeline (Knight Lab, Northwestern)
  • Leaflet - Geographic mapping (open source)
  • TablePress - Bibliographic tables
  • JavaScript - Comparative reading interface

Data Sources

Digitised Copies:

  • Gallica (Bibliothèque nationale de France)
  • Google Books
  • Internet Archive
  • Institutional repositories

Publication Metadata:

  • Penny University Dataset
  • Scholarly articles
  • Gallica, Google Books, Internet Archive, and institutional repositories

Architecture

Website Structure

flowchart TB
    A["Homepage<br/>the-digital-dialectica.rarebook-ubfc.fr/"] --> B["Timeline"]
    A --> C["Publication Map"]
    A --> D["Edition Bibliography"]
    A --> E["Text Comparison"]
    A --> F["About Montesquieu"]
    A --> G["The Dialectic"]
    A --> H["References"]
    A --> I["About"]
    
    E --> E1["Transcriptions"]
    E --> E2["Translation & Semantic Analysis"]
    E --> E3["Comparative Reading Tool"]
    
    I --> I1["Portfolio"]
    I --> I2["Legal Notice"]
    I --> I3["Privacy Policy"]
    I --> I4["Contact"]
    
    B -.uses.-> T1(("TimelineJS"))
    C -.uses.-> T2(("Leaflet Map"))
    
    T1 -.reads.-> D1["Google Sheet"]
    T2 -.reads.-> D2["GeoJSON<br/>(GitHub)"]
    
    D2 -.created by.-> N1["Jupyter Notebook:<br/>Map Data Processing"]
    
    E1 -.created by.-> P1["OCR + Manual<br/>Transcription"]
    P1 -.feeds.-> E2
    E2 -.uses.-> N2["Jupyter Notebook:<br/>LLM API Processing"]
    N2 -.produces.-> D3["Translation JSON<br/>(GitHub)"]
    D3 -.feeds.-> E3
    
    style A fill:#DEFFF8,stroke:#46EDC8
    style B fill:#DEFFF8,stroke:#46EDC8
    style C fill:#DEFFF8,stroke:#46EDC8
    style D fill:#DEFFF8,stroke:#46EDC8
    style E fill:#DEFFF8,stroke:#46EDC8
    style F fill:#DEFFF8,stroke:#46EDC8
    style G fill:#DEFFF8,stroke:#46EDC8
    style H fill:#DEFFF8,stroke:#46EDC8
    style I fill:#DEFFF8,stroke:#46EDC8
    style E1 fill:#DEFFF8,stroke:#46EDC8
    style E2 fill:#DEFFF8,stroke:#46EDC8
    style E3 fill:#DEFFF8,stroke:#46EDC8
    style I1 fill:#DEFFF8,stroke:#46EDC8
    style I2 fill:#DEFFF8,stroke:#46EDC8
    style I3 fill:#DEFFF8,stroke:#46EDC8
    style I4 fill:#DEFFF8,stroke:#46EDC8
    style T1 fill:#E2EBFF,stroke:#374D7C
    style T2 fill:#E2EBFF,stroke:#374D7C
    style D1 fill:#FFDFE5,stroke:#FF5978
    style D2 fill:#FFDFE5,stroke:#FF5978
    style D3 fill:#FFDFE5,stroke:#FF5978
    style N1 fill:#FFEFDB,stroke:#FBB35A
    style N2 fill:#FFEFDB,stroke:#FBB35A
    style P1 fill:#FFDFE5,stroke:#FF5978
Loading

Data Pipeline

  1. Source Materials → Digitised eighteenth-century editions (Gallica, Google Books, Internet Archive, institutional archives)
  2. OCR Processing → Transkribus automated transcription
  3. Manual Correction → Scholarly accuracy checking
  4. Structuring → JSON and CSV formats with metadata
  5. Analysis → LLM-assisted translation and semantic comparison
  6. Validation → Human expert review of computational outputs
  7. Visualisation → Interactive web interfaces
  8. GitHub Repository → Version-controlled open data

What This Project Is (and Is Not)

This project offers:

🔍 Critical examination of digital tools alongside historical texts
📊 Transparency about methodological limits as philosophical findings
🤝 Infrastructure for collaborative scholarship
📚 Humanistic interpretation augmented by computation
🔓 Open scholarship with reproducible methods
⚖️ Epistemological examination of what algorithms built for other purposes such as surveillance capitalism can show about contextual humanistic knowledge
🏛️ Analysis of how digital infrastructure's origins are incompatible with humanistic values

This project does not claim:

✗ That computational methods show objective truth
✗ Techno-utopianism or digital solutionism
✗ That distant reading replaces close reading
✗ That big data is inherently superior
✗ Uncritical acceptance of algorithmic outputs
✗ That digitalisation is ideologically neutral or inevitable

Key Insight: When Montesquieu's philosophy of irreducible contextual complexity meets computational methods requiring abstraction and standardisation, the struggle itself becomes significant data. The limitations show epistemological boundaries and the ideological purposes digital infrastructure serves.


Using This Repository

For Researchers

Clone and examine the data:

git clone https://github.com/lokalkosmos/Lesprit.git
cd Lesprit

All transcriptions, analytical code, and documentation are available for:

  • Verification of scholarly claims
  • Extension to additional editions or chapters
  • Comparative analysis with other Enlightenment texts
  • Critical examination of DH methodologies
  • Investigation of infrastructure politics in digital scholarship

For DH Practitioners

This repository provides:

  • Transparent methodology including acknowledged failures and limitations
  • Reproducible workflows via Jupyter notebooks
  • Critical commentary on tool limitations and infrastructure politics
  • Model for reflexive DH practice that questions its own methods and origins

For Philosophers and Historians

The data enables traditional humanistic enquiry:

  • Close reading of textual variants
  • Translation history analysis
  • Montesquieu reception studies
  • Enlightenment print culture research

The digital infrastructure serves interpretation.


Acknowledgements of Limitations

In the spirit of methodological transparency, this project acknowledges:

Note on Verification: The findings presented require expert verification from specialists in historical linguistics, eighteenth-century philosophy, and textual studies. This verification has not yet been completed.

Corpus Size

  • 15 editions enable depth but limit statistical analysis
  • Trade-off: Textual nuance vs. computational scale
  • Small corpora permit attention to contextual particularity that Montesquieu's philosophy values

Linguistic Expertise

  • Researcher background in philosophy and book history, not historical linguistics
  • Translation analysis relies on scholarly sources and validated automation
  • Honest acknowledgement of expertise boundaries establishes more reliable foundations than false confidence

Computational Opacity

  • LLM operations remain black box despite configuration transparency
  • Cannot verify reasoning behind specific outputs
  • Algorithmic inscrutability contradicts scholarly values of reasoned argumentation and shows systems designed for their intended purposes other than transparent scholarship

Data Structuring

  • JSON formatting imposes uniformity historical texts challenge
  • Material features (typography, marginalia) dissolve into metadata
  • Data preparation enacts the abstraction Montesquieu critiques, showing how computational requirements shape what becomes knowable

Interface Complexity

  • Comparative reading tool creates information density that can overwhelm
  • Multiple windows with lengthy passages may obscure more than clarify
  • Computational presentation fragments philosophical arguments requiring sustained exposition, demonstrating limits of machine-mediated engagement

Geographical Abstraction

  • Map treats culturally distinct cities as equivalent spatial coordinates
  • Geneva ≠ London ≠ Naples in ways location cannot capture
  • Visualisation enables pattern recognition whilst eliminating contextual richness that Montesquieu insists determines meaning

Infrastructure Origins

  • Tools examined emerged from military and corporate research programmes
  • Digital infrastructure designed for command, control, and extraction by design
  • When tools only prove to be inadequate for humanistic understanding, they signal suitability for their intended design and purposes

These are philosophically significant findings about computational epistemology and the ideological purposes digital infrastructure serves.


Citation

Repository

@software{digital_dialectica_2025,
  author = {lokalkosmos},
  title = {The Digital Dialectica and Montesquieu's Book XIX: 
           A Critical Digital Humanities Project},
  year = {2025},
  url = {https://github.com/lokalkosmos/Lesprit},
  institution = {Université de Franche-Comté}
}

Thesis Project's Digital Object

@mastersthesis{2025_montesquieu,
  author = {lokalkosmos},
  title = {The Digital Dialectica and Montesquieu's Book XIX: 
           Critical Tensions in Digital Humanities Scholarship},
  school = {Université de Franche-Comté},
  year = {2025},
  type = {Master's thesis},
  url = {https://the-digital-dialectica.rarebook-ubfc.fr/}
}

Project Website

For citations to specific components:


Licence

This project is licensed under the MIT Licence. See the LICENSE file for details.

All textual transcriptions, data files, and analytical code are freely available for:

  • Academic research and teaching
  • Extension and modification
  • Commercial use (with attribution)

Acknowledgements

Institutional Support

  • Université de Franche-Comté - Hosting infrastructure and academic support

Data Sources

Digitised Copies:

  • Gallica (Bibliothèque nationale de France)
  • Google Books
  • Internet Archive
  • Institutional repositories

Publication Metadata:

  • Penny University Dataset
  • Scholarly articles
  • Gallica, Google Books, Internet Archive, and institutional repositories

Tools and Platforms

  • Transkribus (READ-COOP) - OCR for historical documents
  • Anthropic Claude - Large language model analysis
  • Knight Lab (Northwestern) - TimelineJS
  • Leaflet Contributors - Open-source mapping
  • AI Code Assistance - The development of code for jupyter notebooks and this project's website involved code generation support from AI tools, including Anthropic's Claude 3.5 Sonnet, OpenAI's ChatGPT, and xAI's Grok. These tools were used to assist in drafting and refining code. All code was reviewed, tested, and validated by the author to ensure functionality and accuracy.

Theoretical Foundations

This project builds on critical DH and technology scholarship by:

  • Johanna Drucker (graphical epistemology and rhetorical visualisation)
  • Alan Liu (cultural criticism in DH)
  • Stephen Ramsay (critical digital scholarship)
  • Paul Edwards (infrastructure politics and closed worlds)
  • Shoshana Zuboff (surveillance capitalism critique)

And many others cited in the thesis bibliography.


Contact


Final Note: On Epistemological Modesty

This project concludes where it began: digital tools are useful but fundamentally limited. They make certain patterns visible whilst necessarily obscuring contextual richness. They enable scale but risk sacrificing depth. They promise efficiency but require extensive human labour for scholarly accuracy.

The friction between Montesquieu's contextual humanistic philosophy and computational abstraction illuminates both:

  • What eighteenth-century political thought shows about limits of systematisation
  • What contemporary digital methods show about their own epistemological boundaries and ideological origins

When technical tools built for other purposes deal with texts insisting on irreducible contextual complexity, the forced machine-mediation required reduces the lived experience of contextual difference that humanistic understanding requires. This friction demonstrates tool success at their intended purposes other than nuanced humanistic scholarship.

Approached critically, reflexively, and with epistemological examination of infrastructure's ideological purposes, digital methods can better serve humanistic enquiry into texts that challenge the very operations performed upon them. That friction is where insight begins.


Last Updated: October 2025
Project Status: Active development / Thesis completion

For the most current version of the website and data, visit:
https://the-digital-dialectica.rarebook-ubfc.fr/

About

A critical Digital Humanities project examining Montesquieu's De l'esprit des lois Book XIX through computational methods that expose the friction between contextual humanistic philosophy and digital infrastructure designed for other purposes.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors