Intelligible to Whom?

First published on 2 November 2026 at https://sergiopantojaelt.substack.com/p/intelligible-to-whom

Picture a B2 class. A student is telling the group about her plans.

“My sister wants to leave São Paulo next year. She says the city is too expensive.”

Except leave comes out short, much closer to live.

The teacher doesn’t comment on the pronunciation. She just responds to the content: “Oh really? Where’s she thinking of going?”

A minute later, talking about a holiday, the same student gives beach the same short vowel, which, as most of us know, can land somewhere rather less family-friendly. Again, the conversation just moves on.

For the teacher, there was nothing to repair. Every word got through.

Over the last few years, many of us have stopped chasing native-like pronunciation, and for good reason. The goal now, we often say, is intelligibility. Students don’t need to sound like someone from London or Ohio. They need to be understood.

I totally agree with that. But I’ve been thinking about a question that tends to get skipped.

Intelligible to whom?

When a teacher shares the students’ first language, the honest answer is sometimes “to me”. That is a real answer. It just isn’t the whole one.

Our ear is trained twice

There are at least two reasons we understand our students so easily.

The first is the shared language. If I learned English as a Brazilian, I know where Brazilian speakers’ English tends to go, because I made many of those moves myself. When leave comes out short, I don’t hear a puzzle. I hear the student’s sister packing her bags. Context and experience do the repair before I notice there was anything to repair.

The second is less obvious. We spend our working lives listening to learner English. Accents, half-finished sentences, creative grammar. Our ears have had years of practice that most listeners never get.

So our ear is trained twice: once by the language we share with our students, and once by the job itself.

None of this changes the fact that the student was understood. She was.

The problem begins when we generalize from that single interaction. Just because one listener understood her doesn’t mean others would find her equally easy to understand. And that’s where the illusion lies: treating one person’s understanding as evidence that her speech would be intelligible to a much wider range of listeners.

What the research says

A few distinctions help here.

Munro and Derwing (1995) separated three things we often blur together.

Intelligibility is how much of the message a listener actually understands. Comprehensibility is how easy or hard the listener feels it is to understand. Accentedness is how different the speech sounds from the listener’s own variety. In their study, these were related but partly independent. Speech could be strongly accented and still highly intelligible.

That finding is one of the reasons “sound native” stopped being a sensible goal. An accent, on its own, isn’t the problem.

Jenkins (2000) asked a related question about interactions between non-native speakers with different first languages: which pronunciation features actually seemed to cause breakdowns, and which didn’t? Her proposed Lingua Franca Core is a set of priorities for that kind of communication. It leaves out many features that native-speaker models treat as essential. The th sounds, for example, are outside the core. The contrast between long and short vowels, as in leave and live, is inside it.

This brings us back to the student in my opening example. Jenkins identified the pronunciation feature she struggled with as important for intelligibility. Yet a Brazilian teacher might understand what she meant without even noticing the problem, simply because they’re used to hearing similar pronunciation difficulties.

Are we part of the problem?

To be fair, being an easy listener is part of being a kind teacher. Students need to feel understood before they risk speaking. I’m not suggesting we pretend to be confused.

But a familiar ear can mislead our feedback in two directions.

1- It can let real problems pass. A short vowel in leave might never get flagged, because it never trips us.

2- It can push us to correct things that bother nobody. With our ear doing so much repair, we may fall back on our idea of “correct” to decide what to fix, and spend time on features that rarely get in anyone’s way.

And there’s an important distinction here. This is not an argument for bringing native-like accents back through the side door. For many of our students, the people they’ll speak English with won’t be native speakers, and won’t share their first language either.

The useful question isn’t “Does this sound native?” It’s “Would someone who doesn’t share this student’s first language get the message?”

That is a hard question to answer from the front of a classroom where everyone shares one.

A few things we can try

• Listen as an outsider, on purpose. Every now and then, ask yourself whether a colleague with a different first language would have understood a sentence as easily as you did. If you hesitate, that’s worth a closer look.

• Don’t fill in the meaning too quickly. When a word is ambiguous, try a genuine “Sorry, leave or live?” before you nod. It gives the student a real moment of repair, the kind they’ll need outside class.

• Borrow other ears. With students’ permission, share short recordings with a teacher or learner from a different L1 background, maybe through an online community. Ask them to write down what they heard. Comparing that with what the student meant is often more useful than any correction we could give.

• Choose priorities by impact. When you work on pronunciation, start with the features that actually caused confusion for those outside listeners. Lists like the Lingua Franca Core can help as a starting point, as long as we treat them as hypotheses to check rather than rules.

None of this needs a new syllabus. It mostly needs a slightly different question in our heads while we listen.

Intelligibility was supposed to free our students from an impossible goal. In many ways, it has.

But if one listener is doing all the judging, and that listener shares the student’s first language and has spent years decoding learner English, the goal can become comfortable in a different way.

And perhaps that’s the risk: the better we become at understanding our students, the easier it is to overlook what others might not.