Research using eye tracking shows where the eyes are fixating, rereading or skipping words, and reveals how the brain processes text in real time. (Unsplash)

Published: November 9, 2025 10:17am EST

Authors

1. Victor Kuperman

Professor, Department of Linguistics and Languages, McMaster University

1. Nadia Lana

PhD Candidate, Cognitive Science, McMaster University

1. Olga Parshina

Visiting Assistant Professor of Psychology, Middlebury

Disclosure statement

Victor Kuperman receives funding from NSERC, as Canada Research Chair in Psycholinguistics (Tier 2).

Nadia Lana receives funding from the Social Sciences and Humanities Research Council.

Olga Parshina does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

Partners

McMaster University provides funding as a founding partner of The Conversation CA.

McMaster University provides funding as a member of The Conversation CA-FR.

Middlebury College provides funding as a member of The Conversation US.

View all partners

DOI

https://doi.org/10.64628/AAM.jp9w9qgyt

Reading is a complex cognitive skill that predicts career prospects and social mobility throughout our lifetimes. For newcomers to a country, success often depends on learning to read fluently in a new language.

In fact, language proficiency, including reading fluency, has been found to be the most important factor for successful employment and social participation.

With record numbers of immigrants settling in Canada and migrating globally, understanding how to support reading skill development in a second language is essential.

Writing systems across the world

The foundation of our scientific understanding of the reading process has been narrow, with a majority of studies focusing on reading in English.

But languages are not all written the same way. Some writing systems use letters (like English, Turkish), others use logographs (Chinese, Japanese), syllable characters (Hindi), and more. Some languages are read left to right (Russian, Spanish), and some right to left (Arabic, Hebrew).

Considering how diverse languages are, an interesting question is whether we develop strategies to understand text in our native language and transfer these strategies to additional languages. These are just two of the many research questions that the Multilingual Eye-Movement Corpus (MECO) aims to answer.

What is MECO?

MECO connects researchers from more than 40 countries — including the three of us, the authors of this analysis — to collect eye-tracking data on reading. Eye-tracking uses a camera set-up to record eye movements during reading. It shows where the eyes are fixating, rereading or skipping words, and reveals how the brain processes text in real time.

Participating labs use an identical procedure so that results can be compared across languages. Participants read the same English texts, but each lab then also tests readers in their native language using translated texts, allowing for the data to be compared.

An eye-tracking set-up at a participating lab at McMaster University, Canada. (McMaster University, Humanities), CC BY

One key finding has been that the way someone reads in their first language leaves traces on their second language. In fact, the study reports that approximately half of the variance in eye movement measures in the second language is explained by respective measures in the first language.

For example, writing systems like Korean pack a lot of information into smaller units, and eye-tracking data reflect this: Korean readers skip many words and have shorter eye movements, but make a lot more of them. In a language like Finnish, where words are much longer, information is more distributed and readers tend to spend more time on words and don’t skip them often.

These are strategies that they carry over to their second language, even when the writing system is different.

With 30 languages represented so far (including Korean, Finnish, Greek, Chinese, Dutch, Turkish and Hindi), MECO is the world’s most comprehensive dataset of cross-linguistic eye-tracking data on reading. (MECO), CC BY

MECO has also reported a dissociation between comprehension and eye-movement behaviour. In their second language, readers often achieved similar comprehension scores to native speakers of English, but their eye movements showed more effortful reading (longer fixations, less skipping and more re-readings).

This strategy, the authors note, could be due to the benefits of understanding written materials in an educational or workplace setting outweighing the benefits of speed.

MECO, applied

Language researcher Yaqian Borogjoon Bao joined the network of MECO researchers while studying the cognitive aspects of her native script, traditional Mongolian, at the Chinese Academy of Sciences.

She had this to report in an interview with our team:

“MECO gave me the framework and support to conduct rigorous research. I hope it will inspire others to explore understudied languages and scripts.”

Traditional Mongolian is one of the only writing systems in the world that is read vertically, from top to bottom, instead of horizontally. (Yaqian Borogjoon Bao), CC BY

At the Federal University of Minas Gerais in Brazil, Marina Leite joined MECO as a collaborator while pursuing a degree in Teaching and Education. She told us:

“I hope the MECO data can be used to enlarge the amount of available data about reading in Brazilian Portuguese. The findings could improve education strategies to boost reading comprehension and literacy skills in my country.”

In classrooms where students are balancing multiple languages, research on how native languages affect additional ones can help researchers, educators and policymakers design better strategies for teaching.

MECO aims to fill this gap. All collected data is open access, allowing other researchers to use the data pursue their own research questions about reading.

Want to write?

Write an article and join a growing community of more than 214,200 academics and researchers from 5,349 institutions.

Register now