Lost in Transmission: A Monte Carlo Bipartite Percolation Model of Language Dependence in Traditional Ecological Knowledge Erosion
Princeton Journal of Pre-Collegiate Research · 2026
DOI: 10.67698/pjpcr.2026.0232 ↗Kurz gefasst
Fachartikel in der Zeitschrift Princeton Journal of Pre-Collegiate Research, 2026. Verfasst von Sihyeon Lee.
Abstract (Originalsprache)
Traditional ecological knowledge (TEK) is not lost evenly, and studies usually describe the unevenness by domain, such as medicine, food or ritual. This paper asks how much of that unevenness can be explained by how strongly a knowledge item depends on the community language, and when a domain label adds anything beyond that. We present EthnoDecay, a Monte Carlo simulation that represents a community as a two-sided network of knowledge holders and knowledge items. Aging and migration remove holders; schooling and language shift weaken the links through which items are passed on. All results come from synthetic communities (24 per scenario, about 6,900 items) and are conditional on the assumptions of the simulator. Holder structure together with language dependence predicted 40-year loss with out-of-sample R^2 between 0.59 and 0.92. Language dependence improved prediction even after domain labels were included, by 0.06 under language shift and by 0.30 when young people were still learning, whereas domain labels added nothing once language dependence was known. When language dependence was assigned to items independently of domain, the differences between domains under language shift disappeared, but language dependence predicted loss as well as before. Domain labels became informative only when a separate, domain-specific decline in the occasions of transmission reached about 0.9 to 1 percent per year in the most affected domain. The model has not been tested on field data. Any use of such a ranking should follow the CARE Principles for Indigenous Data Governance (Collective Benefit, Authority to Control, Responsibility, Ethics).
Zitieren
APA
Lee, S. (2026). Lost in Transmission: A Monte Carlo Bipartite Percolation Model of Language Dependence in Traditional Ecological Knowledge Erosion. Princeton Journal of Pre-Collegiate Research. https://doi.org/10.67698/pjpcr.2026.0232
Vancouver
Lee S. Lost in Transmission: A Monte Carlo Bipartite Percolation Model of Language Dependence in Traditional Ecological Knowledge Erosion. Princeton Journal of Pre-Collegiate Research. 2026. doi:10.67698/pjpcr.2026.0232
BibTeX
@article{Lee2026,
author = {Lee, Sihyeon},
title = {Lost in Transmission: A Monte Carlo Bipartite Percolation Model of Language Dependence in Traditional Ecological Knowledge Erosion},
journal = {Princeton Journal of Pre-Collegiate Research},
year = {2026},
doi = {10.67698/pjpcr.2026.0232}
}
EUVIMED