Reaching Readers in Other Languages: Multilingual Research Summaries
A multilingual research summary puts the core of your paper in the languages your readers actually use. This guide covers the evidence on language barriers in science, how to choose languages, where translated summaries can live, and what to check before you publish one.
Say your paper is in English, your fieldwork was in Peru, and the people who could use the findings tomorrow read Spanish. You send the DOI to a district health officer, and the reply is polite and short. The paper reached them. It did not really reach them.
A multilingual research summary is a short, plain-language version of your findings in the languages your readers actually use. It is not a full translation of the paper, and it does not have to be expensive. Here is the evidence on why it matters, how to pick languages, and where translated summaries can live.
The Evidence on Language Barriers
English dominates research publishing, but it is not the only language research is published or read in. In a 2016 PLOS Biology study, Amano and colleagues searched Google Scholar in 16 languages for documents from 2014 using the terms "biodiversity" and "conservation". Of 75,513 documents, 35.6% were not in English, with Spanish, Portuguese, simplified Chinese and French the most common after English.
The same paper points out the reverse problem, which is the one that matters for promoting your own work: as more research is published in English, findings are often unavailable in the local languages of the people who could use them. In its survey of 44 national and regional protected areas in Spain, 54% of the directors who responded (13 of 24) identified languages as a barrier to using scientific papers as an information source for management.
The cost also falls on researchers. A 2023 survey of 908 environmental scientists, also in PLOS Biology, found that among researchers who had published one English-language paper, those from nationalities with moderate English proficiency took a median of 46.6% more time to read an English-language paper than native English speakers, and those from nationalities with low proficiency took 90.8% more. The study also found that non-native English speakers already spend more effort disseminating their research in multiple languages, through non-English abstracts and outreach in two or more languages.
Both studies are in environmental sciences, so treat the numbers as evidence from that field rather than a universal rate. The direction, though, is consistent with the policy picture. The UNESCO Recommendation on Open Science, adopted by 193 countries in 2021, describes open science as making multilingual scientific knowledge openly available, accessible and reusable for everyone. The Helsinki Initiative on Multilingualism in Scholarly Communication asks institutions and funders to make sure researched knowledge is accessible in a variety of languages and that researchers are credited for dissemination beyond academia.
Choosing Your Languages
The goal is not the longest list of languages. It is the languages of the people your paper should reach. Work through these in order:
- Where was the research done? Participants, communities and local practitioners are usually the first audience a translation serves.
- Who acts on the findings? Policymakers, clinicians, teachers or managers in a particular country read in that country's language.
- Who are your co-authors and collaborators? Their networks are the ones most likely to share a summary in their own language.
- Where is your field active? The 2016 study notes that Spanish, Portuguese, Chinese and French together cover the vast majority of non-English scientific documents in its sample, but that is for conservation; your field may differ.
For a worked example, say you have a psychology paper on parenting stress, with data collected in Mexico and Brazil and co-authors in Germany. Spanish and Portuguese come first, because they are the languages of your participants and of the services that support parents there. German comes next, because your co-authors can share it with their networks. Three summaries, each checked by someone fluent, will do more than ten that nobody reviewed.
| Who you want to reach | Language choice | Where the summary goes |
|---|---|---|
| Participants and communities | The language of the study site | Your lab page, partner organizations, a printed handout |
| Practitioners and policymakers | The national language where they work | A one-page brief, linked from your lab page |
| Researchers in your field | Major non-English languages active in your area | Journal supporting information or a repository |
| Collaborators' networks | Your co-authors' languages | Shared by co-authors on their own pages |
Where a Translated Summary Can Live
Some journals give you a slot for it. At the time of writing (October 2026), PLOS ONE's supporting information guidelines list "Alternative Language Abstract" among the common item descriptions, and note that supporting information files are published exactly as provided and are not copyedited, so the accuracy is on you. The 2016 paper and the 2023 paper by Amano and colleagues (both in PLOS Biology) used this route: the 2016 paper includes its lay summary in Spanish, Portuguese, French, simplified Chinese and Japanese.
Repositories are another home. At the time of writing (October 2026), arXiv accepts translations of previously published work, but a translation is a derivative work, so if the original is still under copyright, the translator needs permission from the current copyright holder. arXiv also asks that, if at all possible, the title of a translation include the original title and the original author's name.
Check your publishing agreement before you post a translation of the full paper. A summary you wrote in your own words is usually simpler, but check that too.
Your own page is the home you control. A short section beside the paper on your lab site or faculty profile, with each language clearly labeled and a link back to the paper of record, is easy for co-authors and partners to share. If you have not written the English plain-language version yet, start there; our guide on how to write a plain language summary walks through it, and a translation is only as clear as its source.
A Checklist Before You Publish a Translation
- Write a clean English plain-language summary first, and get your co-authors to agree on it.
- Translate the summary, not the abstract. Abstracts are written for specialists, and translating jargon only moves the jargon.
- Have a fluent speaker who knows the field check every version. Machine translation can draft, but it cannot be the final reader: a summary that misstates a finding in another language is worse than no summary, because you may not be able to spot the error. Even Google's Audio Overviews, which at the time of writing (October 2026) come in 80+ languages, carry a warning that they may contain inaccuracies or audio glitches.
- Keep numbers, units and the direction of every finding identical across languages.
- Label each version with its language, and link every version to the paper of record.
- Check the journal's and publisher's rules if the translation goes anywhere other than your own words on your own page.
Where Amplify Fits
Amplify's audio brief is the part of the order built for this problem. It comes in English, plus 21 more languages on request, 22 in total, including Spanish, Mandarin, German, French, Portuguese, Hindi, Japanese and Korean. It is generated from the PDF you upload and nothing else, and it carries the title, the authors and a link back to the paper of record.
Amplify gives you the controls for that step. Each format arrives as three variants you can edit line by line or regenerate, a person reviews every format before it is released, and nothing publishes until you approve it. If you have a fluent co-author, ask them to read the version in their language before you approve it.
The audio brief comes with the rest of the order: a video abstract (16:9, about 3 minutes), a vertical short (9:16, 45 to 60 seconds) and a one-page, print-ready infographic. The embeddable widget puts every format beside the paper on your own page, and its analytics show reads, listens, questions asked and languages, which is useful evidence for a grant report or a broader-impact section. You keep ownership, we do not train on your work, and you can delete a paper and its artifacts at any time.
Reaching the Readers You Wrote For
Most papers are written in English for good reasons, and that will not change soon. What can change is whether the people your research is about can hear what you found in a language they use every day. Pick the two or three languages that matter for this paper, get each version checked, and put them where your readers already look.
Quick answers
Do language barriers really limit who reads research?
Yes, and they run in both directions. A 2016 PLOS Biology study found that 35.6% of 75,513 biodiversity conservation documents from 2014 on Google Scholar were not in English. In the same paper, 13 of 24 Spanish protected area directors who answered a survey named language as a barrier to using scientific papers for management.
Which languages should I translate my research summary into?
Choose by who you need to reach, not by how many languages you can list. Start with the languages of the communities, practitioners or policymakers your paper is about, then the languages of your co-authors, collaborators and field. Three summaries, each checked by someone fluent, will do more than ten that nobody reviewed.
Can I add a translated abstract to a PLOS ONE article?
Yes at PLOS ONE, at the time of writing (October 2026): its supporting information guidelines list Alternative Language Abstract among the common item descriptions. Those files are published exactly as provided and are not copyedited, so the accuracy is on you. Two PLOS Biology papers by Amano and colleagues, from 2016 and 2023, put translated abstracts or summaries in their supporting information. Check your own journal's guidelines first.
Is machine translation good enough for a research summary?
It can be a starting draft, but have a fluent speaker who knows the field check it before you publish. Even Google warns that its Audio Overviews, which at the time of writing (October 2026) come in 80+ languages, may contain inaccuracies. A summary that misstates a finding in another language is worse than no summary, because you may not be able to spot the error.
