Language as a window into human culture

The study of diverse languages provides a window into human culture. The vocabulary of each language reflects the concepts that are important and shared by a particular group of people. While there are concepts that are commonly expressed across languages, there is significant variation in terms of which concept receives a unique name. For example, English speakers use the words foot and leg to distinguish between two body parts. In contrast, Belhare speakers conflate these two body parts into one word (laŋ). The use of the word laŋ for two different concepts - FOOT and LEG - is called a colexification (François 2008). Patterns of colexifications can reveal universal principles and culture-specific variation. The challenge, however, is to identify the different factors that influence the structure of the lexicon. In our work, we use lexical databases such as the Cross-Linguistic Database of Colexification (CLICS 4) to investigate the connections between concepts within specific semantic domains. We analysed the body part vocabularies of over 1,000 languages and showed that adjacent body parts are more likely to share the same name (Tjuka et al., 2024). However, we also found differences in the way language families structure their body part vocabularies. Apart from frequently occurring body part colexifications such as FOOT-LEG, HAND-ARM, CHIN-JAW, each language family had its distinct system of colexification patterns.  For example, in the Nakh-Dagestan language family, body parts are frequently colexified based on their shared function, as in colexifications between the joints ELBOW and KNEE. In other language families, body parts are often colexified on the basis of their shape, as in ELBOW and HEEL in languages of the Atlantic-Congo family. These variations highlight the need for speakers to acquire culturally relevant perceptual distinctions. The study of colexifications not only deepens our understanding of linguistic diversity but also sheds light on the complex interplay between language, culture, and cognition.

While languages can reveal much about human cultures, cultures might also shape language. According to the “linguistic niche” hypothesis, languages spoken in large societies with many adult second-language learners (societies of strangers) should become grammatically simpler, whereas those spoken in small, close-knit communities should maintain or increase complexity. This hypothesis has become a highly influential explanation for why languages differ in grammatical complexity. It is often cited as evidence that languages adapt to social environments in ways analogous to biological adaptation. However, much of the evidence purportedly supporting this hypothesis relies on cherry-picked case studies or quantitative analyses without appropriate controls for the non-independence of languages. Using the global Grambank database, we examined whether factors such as population size, proportion of non-native speakers, number of neighbouring languages, and language status predict grammatical complexity, while controlling for shared ancestry and geographic proximity. We also distinguished between two dimensions of complexity: the amount of morphology (“fusion”) and how much information must be grammatically encoded (“informativity”). Overall, we found no evidence that sociodemographic factors systematically shape grammatical complexity. Our results cast doubt on the widespread claim that languages reliably simplify or complexify in response to their social environments.