An open research project · Constanța, Romania
Proverbs are how communities remember what works: whom to trust, when to wait, when many hands help — and when too many cooks spoil it. We gathered tens of thousands of them, from a Scottish collection printed in 1663 to sayings still traded today, and mapped where the same wisdom appears across languages, continents, and centuries.
The map: pick any motif and watch its trail across cultures — drag the time slider to replay its documented record through the centuries. The game: thirty seconds on your phone, and the map literally redraws because of you.
An open research pipeline for comparing short traditional texts across languages. It gathers sayings with their provenance, reduces each to a comparable claim, groups equivalents across cultures, and validates every grouping through human judgment with published inter-annotator agreement statistics. The corpus, the source code, and the judgments are openly licensed and citable.
"Where there is smoke, there is fire."
Found in 17 cultures, from Spain to Vietnam, across at least five language families.
Machines can gather proverbs, but only people can judge whether two sayings from different worlds carry the same idea. Every judgment you make is weighed, cross-checked against other players, and folded into the map — and published (anonymously) as open scientific data that researchers anywhere can build on.
Two sayings, side by side, in plain English. Same rule? Same advice? Just the same theme? You decide.
Judgments steer how proverbs cluster across cultures — and audit the algorithms that do it.
Everything becomes citable open data with a DOI. Your nickname stays on the leaderboard; the data stays anonymous.
The oldest voice on the map is a Scottish collection printed in 1663 — most of its sayings still instantly recognizable four centuries later.
…and growing weekly: the corpus crawls open collections, and every new judgment sharpens the picture. Visit the map twice; it will not be the same map.
Pairs like these were recorded continents apart — and verified by human judges as carrying the same idea:
"In wine there is truth."Chinese
"There is truth in wine."Greek
"A stitch in time saves nine."French collection
"A steek in time saves nine."Scottish · printed 1870
"Everyone is creator of his own fortune."Slovak
"Everyone's the blacksmith of their own fate."Hungarian
Every verified link is drawn on the map — and each one started as somebody's thirty-second judgment.
The extraction and comparison method behind the map was presented at the ConsILR 2025 conference on linguistic resources.
The corpus and all anonymized judgments are archived with a DOI on Zenodo, free for any researcher to use.
Every version of the software is on GitHub and archived with a software DOI — anyone can rebuild this map from scratch.
We welcome collaboration with universities, archives, and research groups that hold their own collections of proverbs, idioms, or other short traditional texts.
A partner dataset can be processed in a separate instance of the pipeline, held in its own database and isolated from the public corpus: collection and cleaning, canonicalisation, cross-cultural clustering, and human validation with reported agreement statistics. Partner material is not merged into the open corpus unless the partner asks for it, and the terms of each collaboration are agreed individually.
We are equally glad to hear from groups wishing to contribute data, annotation effort, or expertise in cross-cultural coding, and from colleagues who would simply like to use the method on their own material.
Enquiries are considered as capacity allows. Please write to epelican@univ-ovidius.ro or vladbelciug80@gmail.com.