Charting the Evolution of Early Modern Drama Aesthetics

A Computational Approach

Séminaire – Humanités Numériques Tourangelles

Université de Tours, 04.03.2025

luca giovannini

(Potsdam/Padoue)

  • Computational literary studies
  • CLS and cultural evolution
  • A CLS approach to evolution of early modern drama aesthetics

 

  This presentation: plu.sh/dhtours

ToDAY

1. COMPUTATIONAL LITERARY STUDIES:

A VERY SHORT INTRODUCTION

what are cls, again?

Digital humanities

Digital literary studies

Computational literary studies

According to Jannidis (2020), CLS have three main features:

  1. are conducted with the aid of computational tools and statistical methods
  2. are often carried out from a quantitative perspective (but currently mixed/scalable methods are on the rise)
  3. presuppose models of thought and work derived from the (hard) sciences

Text

Formalisation

Modelling

<body>
  <div type="act">
    <head>Die Erste Abhandelung.</head>
    <div type="configuration">
    <stage>Der Schauplatz lieget voll Leichen-Bilder / Cronen / Zepter / Schwerdter etc. Vber dem Schau-Platz öffnet sich der Himmel / vnter dem Schau-Platz die Helle. Die Ewigkeit kommet von dem Himmel / vnd bleibet auff dem SchauPlatz stehen.</stage>
      <sp who="#ewigkeit">
        <speaker>Ewigkeit.</speaker>
        <l>Die Ihr auff der kummerreichen Welt</l>
        <l>Verschrenckt mit Weh' vnd Ach vnd dürren Todtenbeinen.</l>
        <l>Mich sucht wo alles bricht vnd felt /</l>
        <l>Wo sich Eu'r ichts / in nichts verkehrt / vnd eure Lust in herbes Weinen!</l>
        <l>Ihr Blinden! Ach! wo denckt jhr mich zu finden!</l>
        <l>Die jhr vor mich was brechen muß vnd schwinden /</l>
        <l>Die jhr vor Warheit nichts als falsche Träum' erwischt!</l>
        <l>Vnd bey den Pfützen euch an stat der Quel erfrischt!</l>
        <l>Ein Irrlicht ists was Euch O sterbliche! verführet</l>
        <l>Ein thöricht Rasen das den Sinn berühret.</l>
        <l>Wil jmand Ewig seyn wo man die kurtze Zeit</l>
        <pb n="13"/>
        <l>Die Handvoll Jahre die der Himmel euch nachsiht</l>
        <l>Diß Alter das vergeht in dem es blüht</l>
        <l>In Vnmuth theilt vnd in Vergängligkeit?</l>

CLS PIPELINE

Operationalizing means building a bridge from concepts to measurement, and then to the world. In our case: from the concepts of literary theory, through some form of quantification, to literary texts (Moretti 2013: 13)

key concept: OperaTionaliSING

  • A practical overview: CLS INFRA Survey of Methods in Computational Literary Studies (= Schöch, Dudar, and Fileva, eds. (2023))

  • An attempt at taxonomy (Herrmann et al. 2023):

    • Textometry (Hyberbase, TXM)

    • Stylometry (Delta & Co.)

    • Semantic text extraction (sentiment analysis, topic modelling, word embeddings)

    • CLS ≈ DLS (digital literary stylistics)?

CLS methods

  • First introduced by Allison et al. 2015 in the first Stanford Literary Lab Pamphlet
    • "Formalism, because [we] were interested in the formal conventions of genre; and quantitative, because we were looking for more precise – ideally, measurable – ways to establish generic differences" (6)
  • Link between distant reading approaches and Russian Formalism (Lvoff 2021)
    • focus on textual morphology (Propp)

quantitativE formalism

CLS ECOSySTEM

CLS CORE infrastructures

DRaCOR

2017-

+4k plays

POETREE

2023-

+330k poems

ELTEC

2017-

+2k novels

  • DraCor (Drama Corpora) is an open access platform for research on dramatic literature.
  • Currently: 17 “programmable corpora” in 12 different languages; more than 4000 texts in XML-TEI format.
  • Applications and tools for the CLS, easy API access (e.g. computing network metrics and speaker distribution, SPARQL searches on Linked Open Data, etc.)

Homepage: dracor.org

Paper: Fischer et al. 2019

Tour: bit.ly/einfdrac

CLS CORE INFRASTRUCTURE: DraCor

DraCor: CORPORA

DraCor: TEI-XML ENCODING

DraCor: network analysis

2. CLS and

CULTURAL EVOLUTION

☞ Oleg Sobchuk, "Evolution of Modern Literature and Film", The Oxford Handbook of Cultural Evolution (= Sobchuk 2023)

The core idea of cultural evolution is that cultural change constitutes an evolutionary process that shares fundamental similarities with – but also differs in key ways from – genetic evolution. Humans and other cultural species are the joint product of both our genetic and cultural inheritances. (CES 2025)

 

 

CULTURAL EVOLUTION

  • Literary scholars complain: stories have no genes to be transmitted, cannot breed, survive or get extinct, like biological species
  • Evolution, as a concept, does not require genes, but only three basic components:
  1. a source of variation in a population
  2. a mechanism that prioritizes certain variants over others
  3. a way of transmitting successful variants

CULTURAL EVOLUTION OF LITERAry artifacts

  1. Source of variation → human creativity
  2. Mechanisms of selection
  3. Transmission → history of genre
  • CE is "the evolution of conventions, or techniques, of storytelling: the how-to knowledge passed between the generations of writers and producers"
  • Conventions? Tropes, clichés, genres, archetypes, formulas... forms! (back to the Russians!)

CE of literary artifacts

MECHANISM OF SELECTION

CONTENT-BASED

 

  • hedonic selection: texts with appealing traits are preferred
  • e.g. clues in detective novel (Moretti 2000)

CONTEXT-BASED

 

  • external factors influence the readers' choices: prestige, peer influence, market, etc.

beyond SELECTION

CUMULATION

  • literature evolves through accumulation of innovations over time (increase in complexity)
  • e.g. detective novel (Sobchuk 2018)

COEVOLUTION

  • influence of other coeve cultural phenomena:
    • systems of production
    • political contexts
    • technological breakthroughs

DRIFT

  • random copying: adoption of a trait is driven by chance
  • helps to explain extreme inequality in success distribution

[C]lose reading produces only anecdotal evidence  — suitable for conducting nuanced case studies, but not for testing large-picture theories. [...]

Even the early scholars of literary evolution acknowledged the importance of moving beyond the anecdotal — towards [...] large scale quantitative analysis of trends (Sobchuk 2023, n.p.)

AT last, we need computational

cultural evolution!

3. A CLS approach to drama history

Results from my PhD dissertation:

Evolutive Dynamics in Early Modern European Drama: A Computational Approach 

(Potsdam/Padova, 2021-2024)

research question

How did the different European dramatic literatures develop their own peculiar features during the early modern era?

Corneille, Cinna (1639)

Shirley, The Gentleman of Venice (1639)

THEORETical background

 

 

Franco Moretti

"Modern European literature: A geographical sketch"

New Left Review 

206 (1994), 86-86

 

 

a question of space

Starting assumption: the evolutionary mechanism of literary history is similar to the one taking place in biology:

 

  • New animal/vegetal species are born as a consequence of their movement in new spaces
  • New literary forms are born thanks to the new political-geographical spaces emerging throughout European history

Speciationist model (Darwin's Finches)

          1400                       1500                           1600                         1700                           1800

According to Moretti, the same process of speciation happened in European drama*

🇫🇷

🇩🇪

🇮🇹

🇬🇧

🇪🇸

* he actually speaks of tragedy, but we could try to generalise it

Common model of European drama (based on Seneca and medieval plays)

Discontinuous, fractured, the European space functions as a sort of archipelago of (national) sub-spaces, each of them specializing in one formaI variation.

(Moretti 2013 [1994]: 12)

how to empirically test this theory?

corpus building

  • 150 plays in five languages         (🇮🇹 🇫🇷 🇪🇸 🇩🇪 🇬🇧)

  • time span: 1561-1710 (150 years)

  • purposefully non-canonical approach

Birth locations for the corpus authors (Wikidata)

Text selection and encoding

  1. "Opportunistic" adaptation of texts already available in XML or other digital formats to DraCor standards
  2. Own encoding from scratch for all texts only available as scans (via OCR + ezdrama tool)

pipeline

dracorising/Dockerising

  • Creating offline custom DraCor corpora from local TEI/XML files
  • Exploiting all DraCor tools for CLS analysis
  • Ensuring reproducibility of results

the early modern drama corpus

choosing a METhODOLOGy

  • According to Willand et al. (2017), three main
    methodologies seem to be the most popular options within the field of computational drama analysis:
    • topic modelling of dramatic genres
    • exploration of network structures
    • analysis of character speeches (stylometry).
  • Here: a mix of quantitative formalist approaches
    • content- and language-agnostic
    • form-oriented
  • Boris Yarkho
    • Member of the Moscow Formalist Circle
    • Methodology for a Precise Science of Literature (*2006)
  • Solomon Marcus
    • Mathematical Poetics (1970)
  • Hartmut Ilsemann
    • “Computerized Drama Analysis” (1995)
    • DRAMANALYS.exe

MODELS & FORERUNNERS

"Measurements can be taken of any quantifiable aspect of a text, but figuring out the significance of that metric to an understanding of the text, or better, mapping that metric onto a preexisting critical concept (such as style, plot, or theme), is crucial to making sense of what is being measured" (Algee-Hewitt 2017: 759)

 

Which elements of a play can we operationalise to MODEL IT COMPUTATIONALLY?

MY GOAL: a global approach to dramatic form

The “elementary building blocks” of drama

(according to Kretz 2012: 105)

DIALOGUE

(main space for the transmission of  information)

"Being dialogic is the most immediate feature of a drama" (cf. A. W. Schlegel)

CHARACTER

Understood "not [as] a physical entity, but rather [as] a literary construct" (Pfister)

PLOT

"a connection of events" (Lessing), sjužet (Formalists = the fabula defamiliarised, cf. Shklovsky)

metrics collected

(via the DraCor API or computed independently)

Type Features
Network Size; Average clustering coefficient; Density; Average path length; Average degree; Diameter; Maximum degree; Number of edges; Number of connected components; Ratio, average degree to maximum degree; Ratio, maximum degree to number of characters; Weighted degree distribution; Protagonism; Mediateness
Cast and speech Average characters per scene; Average length of character speech; Speech intensity; Gendered speakers; Collective speakers
Size Number of acts; Number of Segments; Word count, whole text; Word count, spoken text; Word count, stage directions; Number of prose lines; Number of verse lines
Plot All-in index; Final scene size; Drama change rate

what to do with all these statistics?

My approach: building vectors (embeddings) and using them as proxies for the plays themselves

Example_play_1 = {10, 2, 0.5714, 3, 16792}

Example_play_2 = {26, 6, 0.3422, 2, 40098}

num_speakers num_speakers_
groups		 density (SNA) avg_degree (SNA) word_count_sp
Example_play_1 10 2 0.5714 3 16792
Example_play_2 26 6 0.3422 2 40098

experiments with play embeddings

1. DistanceS

Computing Euclidean distances between vectors within 30 year-long timeframes according to two different procedures

Increase in distance = formal diversification (?)

Representing play vectors as points in a coordinate system (via dimensionality reduction methods, e.g. PCA)

play_8

play_5

play_4

play_3

play_2

play_1

play_7

play_6

In this way, it is possible to identify clusters based on formal/structural similarities

2. CLuSterS

SOME WISHFUL THINKING

1561 (similarity)

🇫🇷

🇩🇪

🇬🇧

🇮🇹

🇪🇸

1710 (differentiation)

At the beginning of the timeframe...

... and at its end.

3. patternS

  • Zoom on individual metrics and follow their evolution across the different sub-corpora
  • Compute shifts in absolute value for each metric (i.e., each vector component) across the time span
  • it seems possible to construct a quantitative profile for each different dramatic tradition
  • reproduction experiment on larger DraCor corpora for 🇫🇷 and 🇬🇧 mostly confirms findings

REsults

  • theoretical framework
  • corpus size
  • operationalisation of the concept of drama

Limitations

  • construction of the EmDraCor corpus
  • development of a key methodology (vectorisation based on formal features)

ContributionS

MERCI de votre attention

@lucagiovannini.bsky.social

giovannini@uni-potsdam.de

references

  • Lvoff, B. (2021). "Distant Reading in Russian Formalism and Russian Formalism in Distant Reading". Russian Literature, 122, 29-65.

  • Moretti, F. (1993). "La letteratura europea", in P. Anderson, W. Barberis, e C. Ginzburg, eds., Storia d'Europa, vol. I. Torino: Einaudi, pp. 837–66.

  • Moretti, F. (2000). "The slaughterhouse of literature". MLQ: Modern Language Quarterly, 61(1), 207-227.

  • Moretti, F. (2013). Distant reading. Londra: Verso.

references

references

Made with Slides.com