news · cognition · translation · AI

The same event. Three languages. Three different minds at work.

COGNILANG investigates how the language of publication, translational choices, and discursive framing jointly shape what readers of international news understand, remember, and believe — treating translation as a cognitive process, not a transfer of words, and using AI to objectify interlinguistic transformations at scale.

EN Markets steadied after the ceasefire was announced. LIN-3
FR Les marchés se sont stabilisés après l’annonce du cessez-le-feu. TRA-3
AR استقرت الأسواق بعد الإعلان عن وقف إطلاق النار. PRD-1

One aligned segment from the pilot corpus, with its annotation codes: linguistic LIN, translational TRA, predictability PRD.

in one paragraph

What the project does

An event breaks. Within hours it is written, rewritten, and translated across the world's press. These versions differ — in framing, in information order, in what is made explicit and what is left out — and those differences are not cosmetic: they change the cognitive work a reader must do. COGNILANG builds a comparable corpus of international news in English, French, and Arabic, annotates it on linguistic, discursive, translational, and predictability variables, uses AI for alignment and divergence detection, and tests the cognitive consequences (comprehension, recall, cognitive load, perceived credibility) with real readers. The protocol is designed to be replicated by other teams in other language pairs.

H1–H2

Texts differ measurably

Framing, density, and modalisation vary by language; translation leaves observable traces — omission, explicitation, modulation, reorganisation.

H3–H4

Readers feel the difference

Those transformations affect comprehension, recall of key information, and credibility — and originals are not processed like translations.

H5–H6

AI and prediction connect them

AI detects transformation regularities at scale, and linguistic prediction (surprisal) links textual predictability to processing effort.

why join

Built to travel across languages

A fixed scientific core

The constructs, hypotheses, annotation grid, validation rules, and analysis logic are invariant — so results from a Spanish–Chinese team are commensurable with the EN/FR/AR pilot.

A documented adaptation layer

Tokenisation, normalisation, framing-label piloting, and language-model choice are explicitly language-specific, with written rules for how to adapt them and report what you changed.

Shared instruments

An 18-worksheet researcher pack, an AI-assisted analyzer app, code-books with examples, and a six-week workshop plan — everything a new team needs to start.

A growing comparative dataset

Each language pair adds a panel to the same picture: how mediation across languages reshapes the world's news, and what that does to its readers.