We often say that some languages have “more words” for certain things, but that claim tends to collapse under scrutiny. Either it is treated as obvious, or dismissed as romanticised nonsense. The real question has never gone away: how uneven is vocabulary across languages, & what drives those differences?
Linguists refer to this unevenness as lexical elaboration: the extent to which a language builds a dense cluster of words around a particular concept, distinguishing subtypes, practices & meanings rather than relying on a single broad term. A new study takes that intuition seriously & asks a simple but powerful question: can we actually measure lexical elaboration across hundreds of languages, at scale, without relying on anecdotes?
The study
Khishigsuren, Regier, Vylomova & Kemp analysed lexical elaboration using a large computational approach. Instead of eliciting word lists from speakers, they built a dataset called BILA (Bilingual Lexicon Assembly), drawing on 1,574 bilingual dictionaries covering 616 languages.
Their key move was methodological. Rather than counting “words” directly, they used the frequency of English glosses in dictionary definitions as a proxy for how densely a concept is elaborated in a language. For example, if snow appears repeatedly across many definitions in an Inuktitut (the language spoken by the Inuit) dictionary, that suggests a large network of snow-related lexemes [1].
They normalised for dictionary size, worked some statistical magic & linked linguistic patterns to ecological variables like temperature, rainfall, windspeed & cultural factors such as subsistence type.
Crucially, they first validated the method by testing 163 well-known claims from the literature about lexical elaboration.
The findings
147 of the 163 previously proposed claims were supported by the data, & 63 ranked above the 95th percentile for strength of evidence. Taken together, the results show that lexical elaboration is neither random nor purely environmental: it reflects a mix of ecology, culture & communicative priorities.
Some highlights:
- SNOW: The much-mocked claim about Inuit languages & snow was strongly supported when examined comparatively. Crucially, this emerged from large-scale dictionary data across hundreds of languages, not from isolated examples, giving the claim empirical credibility.
- Climate terms: Environmental concepts showed systematic patterns. Lexical elaboration for SNOW & ICE decreased as average temperatures rose, while WIND-related vocabulary increased with windspeed, indicating that persistent environmental pressures leave consistent lexical traces.
- RAIN: The pattern broke down for rain. High rainfall alone did not predict rich rain vocabularies. In several drier regions, rain was lexically elaborated because of its cultural value, symbolic role or social practices tied to scarcity & anticipation.
- SMELL: Sensory domains mattered. SMELL showed strong lexical elaboration in Oceanic & South Asian languages, extending earlier work by Majid & others and suggesting that olfactory salience is culturally distributed, not biologically uniform.
- DANCE: Cultural concepts behaved differently again. DANCE showed high lexical elaboration in smaller societies, pointing to the role of social organisation, ritual life & communal practice rather than environmental necessity.
In other words: a language might develop many words for wind not simply because it is windy, but because wind affects fishing, navigation or ritual life. Likewise, a community might talk extensively about rain precisely because it is rare.
Why this matters
This work connects neatly with long-standing ideas from Nida about cultural salience, with communicative need accounts in lexical semantics, & with usage-based views of language that see frequency & function as shaping form.
For ELT professionals, there’s no simple takeaway, but it’s a reminder that vocabulary size in a given domain is not random, nor simply a reflection of “complexity”. It reflects what speakers notice, need to talk about & care about repeatedly.
Consider how English elaborates time management or transport, while learners’ L1s may elaborate kinship, taste or landscape instead. These are not gaps to be “fixed”, but different mappings of experience onto language.
Teacher takeaways?
- Be cautious with textbook assumptions about “basic” vs “specialised” vocabulary. What feels niche in English may be central elsewhere.
- When teaching lexis, foreground why certain distinctions exist, not just what they are. Salience is cultural as well as cognitive.
- Use contrastive discussion productively: ask learners where their own languages are more precise than English, & why.
Are vocabulary differences deficits, or evidence of different ways of carving up experience?
[1] a basic unit of meaning in a language, abstracted away from its inflected forms. For example, run, runs, ran & running all belong to the same lexeme, while run (the verb) & run (a tear in fabric) are separate lexemes because they encode different meanings.
Leave a Reply