A typology of lexical structures


The marama project – led by Alexandre François (CNRS, Paris; ANU) and Siva Kalyan (Univ. of Queensland, Australia) – includes a particular interest for lexical structures across the world's languages, i.e. the manner in which meaning is distributed across linguistic forms in lexicons of the world. This research agenda is known as lexical typology.

From etymology to lexicography

François developed his passion for lexical semantics as he studied classical languages (Greek, Latin, PIE reconstruction...) and learned about semantic change from ancient languages to modern times. His interest was renewed as he researched the Oceanic languages of Vanuatu: this has led to various lexicographic projects, in the form of printed or electronic dictionaries on various languages of Vanuatu (Araki, Mwotlap, Lo-Toga, Hiw) or the Solomons (Teanu, Lovono, Tanema).

Colexification and the typology of synchronic polysemies

map-Banks In parallel with his descriptive work on Oceanic languages, François began searching for universal patterns of polysemy by proposing the new concept of colexification – that is, the phenomenon whereby distinct concepts can be expressed identically (=can be “co-lexified”) in certain languages.

François' concept of colexification – and/or the method he proposed for building semantic maps to represent patterns of colexification – have since inspired numerous scientific works, in the domains of lexical typology, cognitive semantics or natural language processing (NLP).

Semantic change and “lexical tectonics”

More recently, François endeavoured to broaden his reflection by focusing on the diachronic dimension of semantic associations, namely semantic change and etymology. This took the form of various publications:

The metaphor of “lexical tectonics” captures the phenomenon of semantic change in the lexicon, as word meanings progressively evolve over time. The slow reshaping of each word's profile in semantic space somehow reminds us tectonic plates drift in a way that slowly reshapes our geographical landscapes.

Dialexification, a tool for studying semantic change

In 2023, in order to study the diachronic dimension of colexification, François and Kalyan coined the new concept of dialexification. Dialexification captures the relation between “cognate meanings”: that is, if two words in different languages are cognate (=descended from the same etymon), then the meanings of those words will be dialexified. For example, French entendre 'hear' & Spanish entender 'understand' show that the two concepts {hear—understand} are “dialexified” in the Romance family.

𝓔𝓿𝓸Sem: A database of dialexification

This concept of dialexification has given rise to a new online project: 𝓔𝓿𝓸Sem. This database of dialexification explores historical relations between concepts, based on their etymological links in dozens of different language families across the world. Evosem

Alexandre François, Siva Kalyan, Mathieu Dehouck, Martial Pastor & David Kletz. () 𝓔𝓿𝓸Sem: A database of dialexification across language families. Online database. CNRS—LaTTiCe, Paris. https://tiny.cc/EvoSem.

Visit the 𝓔𝓿𝓸Sem database.