In this article, we analyze the temporal and geographic evolution of sustainability-related discourses over a time frame of twenty years (1999-2018). We use a collection of multilingual newspapers in English, French, German, Spanish, and Italian, as a proxy. We filter documents using four key terms: sustainable development, climate change, environment, and pollution, seeking to explore how different newspapers encode the same message, aiming to detect points of contact (agreement) and rupture (polarity). Our methodology includes Topic Modelling (Pachinko Allocation [1]), word embeddings [2], Ward’s hierarchical cluster analysis [3], and network analysis [4]. Our results show a progressive simplification of semantic fields over time, reflecting less polarizing views across countries and, therefore, showing an increasing agreement on sustainability-related discourses in our contemporary societies. Moreover, we also notice little variation of newspapers rhetorics over time. Therefore, this article also contributes with a meta-reflection about newspapers behaviour as information containers.