Human language is not a fixed set of rules. It is an organic, adaptive system that continuously evolves through time and social interaction. Words are born, their meanings expand or contract, sometimes shift entirely, and occasionally disappear. In linguistics, this phenomenon is known as lexical semantic change or semantic shift, and it serves as a mirror reflecting cultural, political, and technological transformations within a society.

For example, the word mouse once referred exclusively to a small rodent. With technological change, its meaning expanded to denote a computer input device. More recently, words such as woke, which once conveyed positive social awareness, have been repurposed within politically polarized discourse and are now often used with negative connotations.

Traditionally, the study of semantic change relied on the intuition of philologists and manual diachronic corpus analysis. While valuable, this approach struggles to capture subtle or rapid shifts within today’s overwhelming volume of textual data. As access to large scale historical corpora has expanded and natural language processing techniques have advanced, the study of meaning change has undergone a fundamental transformation. Researchers can now quantitatively trace semantic trajectories across centuries of literature as well as real time streams of social media text.

This article introduces recent research that analyzes semantic change using two complementary methodological pillars: diachronic word embeddings and text network analysis. Used independently or in combination, these approaches have been applied to interpret social phenomena, digital subcultures, and business ecosystems.

Read more >>