Narrative Structure in Tropes - A Computational Analysis of 'Friends'

Aug 24, 2026

Upload by Riley

Recurring narrative patterns known as tropes are recurring devices in television and film. This study performs a computational analysis of tropes in the sitcom Friends using human-curated trope annotations from TVTropes, episode transcripts, and IMDb ratings. Rather than attempting direct automatic detection, it treats existing annotations as a curated analytical layer. It finds a statistically significant positive association between trope frequency and weighted IMDb ratings, then links trope annotations to dialogue semantics via TF-IDF features, clustering 1,954 tropes into 15 interpretable groups and examining their character and semantic organization.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages18

Reading Time45 min

AccessFree to download

Introduction

Tropes as recurring narrative devices

Prior research and challenges in automatic detection

Data source and study design: TVTropes and Friends

arXiv :2606 . 19499v1 [physics .soc-ph] 17 Jun 2026 Narrative Structure in Tropes: A Computational Analysis of ‘Friends’ Shun Zhang, 1, ∗ Tabia Tanzin Prama, 1 Christopher M. Danforth, 1, 2,† and Peter Sheridan Dodds 1, 3, 4, 5,‡ 1 Computational Story Lab, Vermont Complex Systems Institute, MassMutual Center of Excellence for Complex Systems and Data Science, Vermont Advanced Computing Center, University of Vermont, Burlington, VT 05405, USA. 2 Department of Mathematics & Statistics, University of Vermont, Burlington, VT 05405, USA. 3 Department of Computer Science, University of Vermont, Burlington, VT 05405, USA. 4 Santa Fe Institute, 1399 Hyde Park Rd, Santa Fe, NM 87501, USA. 5 Complexity Science Hub, Metternichgasse 8, 1030 Vienna, Austria (Dated: June 19, 2026) Tropes are recurring narrative devices in television and film. We carry out a computational analysis of tropes in the sitcom Friends, using human-curated trope annotations from ‘TVTropes’, episode transcripts, and IMDb ratings. Because automatic trope detection remains challenging, we treat existing trope annotations as a curated analytical layer and focus on their downstream narrative and semantic functions. We first examine the relationship between episode-level trope frequency and audience reception. We find a statistically significant positive association between trope count and weighted IMDb ratings, although the modest explanatory power suggests that more than just trope density explains audience evaluation. We then connect trope annotations to dialogue transcripts and represent trope-related dialogue using TF-IDF-based semantic features. Using PCA and k-means clustering, we group 1,954 distinct tropes into 15 semantically interpretable clusters. Chi-square analyses show that the six main characters are unevenly distributed across these clusters, with character-specific trope profiles that are broadly consistent with their established narrative identities. Finally, we project trope clusters into the ousiometric power–danger space to examine their semantic organization. The results show that ‘Physical and Sexual Comedy’ occupies a region associated with relatively high danger, while ‘Revelation, Surprise, and Reaction’ occupies a region associated with relatively high power. Overall, our work first demonstrates a way to operationalize trope measurement, and second, shows that identifiable trope clusters can provide holistic ‘distantreading’ descriptions of characters and stories. I. INTRODUCTION Recurring narrative patterns constitute a fundamental component of storytelling of all kinds. Originally framedas motifs for folktales, such conventions for television and film came to be commonly referred to as tropes [1, 2] . Tropes may be classic plots, short sequences of actions, jokes, or types of character interactions. Tropes are important because human beings view the world through structured narratives and accumulated experience [3] . These narrative structures recur across a vast number of works, thereby shaping viewers’ expectations regarding character and plot development [4] . Previous research has demonstrated that large collections of narratives consistently exhibit recurring structural patterns [5, 6] . Many stories rely heavily on such repeated narrative frameworks [7], and tropes are among them. This leads us to our aim of quantitatively examining the structural functions of tropes in storytelling. Compared with themes or broad genres, tropes are more concrete but still more general than individual plot events. This intermediate level of abstraction makes them particularly suitable for our systematic analysis. ∗ † ‡ [shun.zhang@uvm.edu](shun.zhang@uvm.edu)[ ](shun.zhang@uvm.edu)[chris.danforth@uvm.edu](chris.danforth@uvm.edu)[ ](chris.danforth@uvm.edu)[peter.dodds@uvm.edu](peter.dodds@uvm.edu) Here, we use the “Brick Joke” trope as an example to illustrate the function of tropes as recognizable and repeatable narrati

FAQs

How does the analysis operationalize tropes without relying on automatic trope detection?

The study uses human-curated trope annotations from TVTropes as a stable ground truth, focusing on downstream narrative and semantic functions rather than direct detection from content.

What relationship is found between trope frequency and audience reception?

An episode-level trope count shows a statistically significant positive association with weighted IMDb ratings, though the explanatory power is modest.

How are the tropes organized for interpretation in the study?

The work represents trope-related dialogue with TF-IDF-based semantic features, then applies PCA and k-means clustering to group 1,954 distinct tropes into 15 semantically interpretable clusters.

Aug 24, 2026

Upload by Riley