Follow the leader: Documents on the leading edge of semantic change get more citations

Abstract

Diachronic word embeddings—vector representations of words over time—offer remarkable insights into the evolution of language and provide a tool for quantifying sociocultural change from text documents. Prior work has used such embeddings to identify shifts in the meaning of individual words. However, simply knowing that a word has changed in meaning is insufficient to identify the instances of word usage that convey the historical meaning or the newer meaning. In this study, we link diachronic word embeddings to documents, by situating those documents as leaders or laggards with respect to ongoing semantic changes. Specifically, we propose a novel method to quantify the degree of semantic progressiveness in each word usage, and then show how these usages can be aggregated to obtain scores for each document. We analyze two large collections of documents, representing legal opinions and scientific articles. Documents that are scored as semantically progressive receive a larger number of citations, indicating that they are especially influential. Our work thus provides a new technique for identifying lexical semantic leaders and demonstrates a new link between progressive use of language and influence in a citation network.

Abolitionist Networks: Modeling Language Change in Nineteenth-Century Activist Newspapers

“…In the digital humanities, diachronic word embeddings have been used to trace the history of concepts: for example, Shechtman [2020] explores the "technical, ideological, and environmental valences" of the idea of media by comparing near neighbors of media and related terms across a corpus of 20th century magazines. Most closely related to our own work, Soni et al [2020] present a method for identifying the specific documents that lead lexical semantic changes, which the authors show to accrue more citations in corpora of scientific abstracts and legal opinions. This work provides further evidence for the validity of our approach of linking lexical semantic leadership to broader influence.…”

Tracking Semantic Shifts in German Court Decisions with Diachronic Word Embeddings

“…While multiple works have been published training diachronic word embeddings on political data and debates, including the previously mentioned 2021), but also work from Rozado and Al-Gharbi (2022) an Indukaev (2021), there is very little work focusing on the legal domain, although the idea that training diachronic distributional models on legal language could provide valuable insights has been voiced before (Rice, 2019). Soni et al (2021) were the first, and, as far as we are aware, so far the only ones, to specifically train diachronic word embeddings on court decisions. By training on decisions from US federal courts, they were able to identify decisions that are "on the leading edge of semantic change" and show that such decisions are cited more often.…”

Naukometriya, Nalimov and Mul’chenko

“…As a start, they point out that natural languages are polymorph, which means that they do not have strictly defined boundaries and that depending on the context a word may have different meanings. Moreover, meanings of terms may change over time [42]. The authors then note that the language of scientific citations is even more polymorph: a single bibliographic citation encompasses many ideas.…”