When a multinational team runs every meeting in English, the cost isn't usually a breakdown — people manage. The cost is quieter: the same understanding takes more mental effort for non-native speakers, the quietest people in the room often aren't the ones with the least to say, and the nuance that gets lost is invisible because nobody flags what they didn't quite catch. None of that shows up in the meeting notes. But it's real, and a fair amount of it has been measured. Here's what the research shows — and why I think it's the strongest argument for translating meetings rather than defaulting to a shared second language.
Understanding costs more, even when comprehension looks fine
The most striking finding is that a non-native speaker can understand exactly as much as a native speaker and still pay more for it. In a 2018 pupillometry study, Borghini and Hazan measured listening effort by pupil dilation and found it was significantly higher for non-native listeners even when both groups achieved equivalent comprehension accuracy. The understanding was the same; the cognitive cost was not. In a meeting, that extra effort is spent on simply keeping up — leaving less for analysis, pushback, or remembering to make your own point.
A second language changes how people decide
It's not only effort. Operating in a foreign language seems to change reasoning itself. Keysar and colleagues' work on the "foreign-language effect" found that classic decision biases like framing and loss aversion shrink when problems are posed in a non-native language, because a foreign tongue is processed with more deliberation and less automatic, intuitive fluency. That cuts both ways — sometimes more deliberation is good — but the point stands: people are not the same thinkers in their second language as in their first. A team that only ever hears its members in English is hearing a slightly different version of them.
Quiet isn't agreement
Anyone who runs cross-language meetings has seen it: some people just don't talk much. It's easy to read that as having nothing to add. The research suggests otherwise. Rogerson-Revell's study of international business meetings found that native English speakers didn't dominate talk time, yet there was "a much higher proportion of inactive non-native English speakers" in the room. The silence wasn't about content; it was about the cost and confidence of speaking up in a second language. Decisions made from the voices that spoke are missing the ones that didn't.
The time tax is large where it's been measured
Direct measurement of meeting costs is hard, but an adjacent domain gives a sense of scale. A 2023 study in PLOS Biology on non-native English speakers in science — a reasonable proxy for high-stakes knowledge work — found they spent a median of 46.6% more time reading an English paper, 50.6% more writing one, and 93.7% more time preparing an oral presentation in English, with papers rejected for language reasons over twice as often. These are publishing figures, not meeting figures, so treat them as an analogue rather than a direct measurement. But the direction is unambiguous: doing professional work in a second language carries a substantial, repeated time penalty.
It's a management cost, not just a personal one
This rolls up to the organization. In her Harvard Business Review piece on English as a corporate language, Tsedal Neeley documented how mandated-English policies at global companies produced real friction — and how highly competent employees can feel reduced, even "childlike," when forced to operate in the lingua franca. A more recent systematic review of 122 papers on language-related misunderstanding at work links it to measurable harm to performance, commitment, and team trust. When your best engineer or operator goes quiet in a second language, that's not a soft cost.
The Asian-first angle
This is sharpest for teams spanning Asian languages, where the distance from English is largest and the "everyone just speaks English" default is most quietly expensive — a point I've argued separately in why "everyone just speaks English" is failing your APAC team. When a Tokyo or Jakarta or Taipei colleague has to operate in English while their counterparts in London operate in their first language, the effort tax isn't shared equally — it lands on the people furthest from the lingua franca.
What to do about it
You don't fix this by telling people to "just speak up." The effort and the participation gap are structural, not attitudinal. What helps is removing the requirement to operate in a second language in the first place — letting each person speak and read in their own. That's the case for real-time meeting translation: not a luxury, but a way to get the full team's thinking into the room instead of only the part that survives translation in people's heads.
That's what we built Sageio to do — each participant reads meeting captions in their own language (20+ languages, built Asian-first), with a translated transcript and summary afterward, so the record is usable by everyone. If the costs above sound familiar, that's the lever.
Frequently asked questions
Is there real evidence that working in a second language has a cost? Yes. A 2018 pupillometry study found non-native listeners exert more cognitive effort even at equal comprehension; a 2023 PLOS Biology study found large time penalties for doing professional work in English as a second language; and research on international meetings found non-native speakers participate less. The cost is measurable, not just anecdotal.
Does speaking a second language actually change how people think? Research on the "foreign-language effect" found that decision biases like framing and loss aversion shrink in a non-native language, because it's processed more deliberately and less automatically. People reason somewhat differently in their second language than their first.
Why is this worse for Asian-language teams? Because the linguistic distance from English is larger, so the effort cost and the participation gap tend to be greater — and the "everyone speaks English" default disproportionately taxes the colleagues furthest from English while their English-first counterparts pay nothing.
How does real-time translation help? It removes the requirement to operate in a second language during the meeting — each person speaks and reads in their own — so the effort tax and the quiet-non-native-speaker problem are reduced at the source, and the full team's input makes it into the room and the record.
"Everyone speaks English" usually works well enough to never get questioned. The research is a good reason to question it: the cost is real, measurable, and unevenly distributed. If your meetings span languages — especially Asian ones — letting people work in their own is worth a real test. See how it works.