Native-Speakerism in Pronunciation Learning: Choose a Model Without Ranking People

Choose an English pronunciation model without ranking people: match the reference to your goal, listeners, identity, access, and practice needs.

Sofia Laurent

FunFluen editor · Idioms and culture

Aug 31, 2026 9 min

You can choose General American, General British, or a specific native-community speaker as your pronunciation model without ranking the rest of humanity. The problem is not the reference. The problem starts when native becomes shorthand for best, clearest, most correct, or most qualified.

Use this model-choice filter: REFERENCE → JOB → FIT → BOUNDARY. Or, in one line: choose a pronunciation reference, not a human hierarchy.

What native-speakerism actually means

Adrian Holliday introduced native-speakerism in English-language teaching as an ideology that gives “native speakers” and an imagined Western cultural/teaching ideal superior status while assigning inferior roles or qualities to “non-native speakers.” Read Holliday’s foundational ELT Journal article.

That definition matters because it stops the term from becoming uselessly broad. Native-speakerism is not:

• noticing that someone grew up speaking English;
• using a native-community recording as one reference;
• choosing General American or General British for consistency;
• personally wanting a very native-like performance.

The hierarchy appears when native status is allowed to do work it cannot logically do by itself: “therefore this person is the best teacher,” “therefore this accent is the correct English,” “therefore this speaker is clearer,” or “therefore everyone else should move toward this voice.”

MODEL SOURCE ≠ MODEL QUALITY ≠ HUMAN WORTH.

“Copy a native speaker” skips the important questions

Imagine someone tells you, “Pick a native speaker and imitate them.” Fine—but which one?

A Glasgow speaker, a Texan, a Londoner, a Jamaican, a New Zealander, a Singaporean native English user, and a broadcaster in Toronto can all be native English speakers and still give you very different pronunciation targets. Even labels such as General American or General British are reference categories, not recordings of an entire country speaking with one mouth.

So native gives you background information. It does not finish the model-selection job interview.

Nativeness and intelligibility are different goals

Pronunciation research has long contrasted what John Levis calls the Nativeness Principle with the Intelligibility Principle. The first treats native or near-native pronunciation as the target; the second prioritizes speech that works for communication even when an accent remains noticeable.

In a 2024 article explicitly challenging native-speakerism in pronunciation teaching, Levis argues that native accents should not be treated as universal desired outcomes and that intelligibility gives a better basis for deciding which pronunciation features deserve teaching attention. Read Levis’s 2024 article on the Intelligibility Principle.

That does not mean a learner is forbidden to want native-like pronunciation. If your informed, voluntary goal is to sound close to a specific community model, that is a legitimate personal or performance goal. It simply is not the communication benchmark everyone else must share.

Do not replace one hierarchy with another

It would be wonderfully efficient—and wrong—to solve native-speakerism by declaring, “Fine. The most intelligible speaker is always the best model.” That just builds a new ladder.

Intelligibility depends partly on the speech signal, but also on listeners, familiarity, context, expectations, and interaction. A model that works beautifully for one learner’s community may not be the right production target for someone preparing for a specific local workplace, acting role, family community, or course.

A 2025 System review goes further: it argues that ideologies of nativeness can remain hidden even inside supposedly intelligibility-focused pronunciation research, especially when privileged L1 varieties remain the silent norm. The authors recommend broader perceptual repertoires and caution against turning prescribed pronunciation models into universal targets. Read the 2025 review on ideologies of nativeness.

So clarity matters. Listener fit matters. But “the clearest voice wins forever” is not the replacement philosophy.

Nonnative and multilingual speakers can be useful models too

Pronunciation-model research has explicitly explored intelligible, comprehensible nonnative English speech as a teaching model. In a 2014 System article, John Murphy argues that intelligible and comprehensible nonnative speech can offer useful aspirational and relevant models for ESL/EFL learners; the article also reports a questionnaire study in which 34 pronunciation specialists evaluated one nonnative English speech sample. Read the article on intelligible nonnative pronunciation models.

That does not prove that every nonnative speaker is a good model—just as native status never proved every native speaker was. The selection rule stays the same: model quality has to be demonstrated by fit.

Recent pronunciation-teaching research also shows that the native-speaker ideal is not fixed or self-evident. A 2024 open-access study by Janine Berns and Sanne van Vuuren examines how ideas about “good pronunciation” and native-speaker norms are evolving in English and French teaching practice. Read the study on evolving native-speaker ideals in pronunciation teaching.

REFERENCE: make your model exact

“I want a native English accent” is too vague to practise.

Better references look like:

• “I’m using this General American dictionary/audio model for isolated word stress and /r/.”
• “I want the pronunciation used by this particular British English teacher for my course.”
• “I’m modelling speakers from the community I’m moving into.”
• “I’m using this highly intelligible multilingual English speaker because their pacing and phrasing match my international workplace.”
• “I’m using this actor/coach reference for one role.”

A reference is useful when it is specific enough to compare against. “Native” is a category; practice needs an actual target.

JOB: what are you hiring the model to do?

Different goals can justify different models.

A model should earn the job. “Native” is not a complete résumé.

FIT: use the Model Fit worksheet—no nativeness score

This is a practical decision framework, not a scientific ranking system. Compare one to three candidate references.

Notice what is missing: there is no row for “percentage native.” If two models both fit, you do not need a philosophical winner—pick the one you can practise consistently.

Your final decision can be one sentence:

“For production, I choose ___ because ___; my main listeners are ___; I will keep listening to ___ as well.”

BOUNDARY: write what your choice does not prove

This is the step that keeps a practical reference from becoming a human hierarchy.

Finish this sentence:

“I choose this model for ___; this does not mean ___ is inferior.”

Examples:

• “I choose General American because most of my learning input is U.S. media; this does not mean Nigerian, Indian, Scottish, or other Englishes are wrong.”
• “I choose a local native-community model because I’m moving there; this does not make a native speaker automatically a better teacher.”
• “I choose an intelligible nonnative model because it fits my international team; this does not mean native-community speech is bad.”

The opposite of native-speakerism is not “native speakers are bad.” Congratulations—we would have built the same ladder upside down.

Try the cases before opening the answer

One production model, broad listening

There is a useful asymmetry here: production can benefit from a stable reference, while listening often benefits from wider exposure.

The 2025 System review specifically recommends broadening learners’ perceptual repertoires rather than keeping privileged L1 varieties as the silent default. You do not need to pronounce every variety you can understand. You do need enough listening breadth for the Englishes you will actually meet.

A simple plan:

• Production: one main reference for consistency.
• Listening: multiple speakers, regions, L1 backgrounds, styles, and contexts relevant to your life.
• Boundary: never confuse “my production target” with “the correct English.”

If you choose a model to practise

Once the model has earned its place through REFERENCE → JOB → FIT → BOUNDARY, practise it because you chose it for a reason. FunFluen can support general English speaking practice after that decision. It does not rank English varieties, speakers, or native status.

Practise the model you chose for a reason

One reference, many legitimate Englishes

You do not need a pronunciation democracy where every practice sentence uses a different model. Consistency can be useful. You also do not need a pronunciation monarchy where one ancestry category gets permanent authority over everybody else.

Pick the reference. Name the job. Check the fit. Write the boundary.

Choose a pronunciation reference, not a human hierarchy.

Sources

• Holliday — Native-speakerism
• Levis — Challenging Native-Speakerism: Embracing the Intelligibility Principle in Pronunciation and Spoken Language Instruction
• Beyond ideologies of nativeness in the intelligibility principle for L2 English pronunciation
• Murphy — Intelligible, comprehensible, non-native models in ESL/EFL pronunciation teaching
• Berns & van Vuuren — The native speaker ideal in FL pronunciation teaching practice

Explore more language-learning guides in Media-Based Language Learning.