EXAMINING DRIVERS OF LINGUISTIC ENDANGERMENT USING MAXIMUM ENTROPY MODELING

Publication Type	Journal Article
Publication Year	2026
Author(s)	Maeve Malone and Jennifer Santoro
Journal Name	The Pennsylvania Geographer
Volume, Issue	63, 2
Pagination	1-16
Article Type	Paper
DOI	https://doi.org/10.66518/UPFW1993

Keywords: endangered languages, maximum entropy modeling, linguistic diversity, species richness

Article: DOWNLOAD

Abstract: Global language diversity is decreasing at a rapid rate: almost 40% of the world’s

7,100 languages are at risk of becoming extinct in the next century. When languages

die, indigenous knowledge and cultures die with them. Analyzing various spatial and

environmental factors that are shared between these endangered languages could help

prevent their decline and identify languages that are at a greater risk of becoming

endangered. This paper examines 2,715 languages identified by UNESCO across five

endangerment classifications from vulnerable to extinct. Analysis using Maxent indicates

that ecological factors including species richness and human factors including

population density may influence the probability of language endangerment, indicating

that biodiversity may play an important role in predicting the probability of language

endangerment and extinction in the future. Using Maxent to model language

endangerment represents a novel method to study these patterns and develop solutions

for language and ecological conservation.