Language & Intelligence

Why your recorded voice sounds unfamiliar, and how expectations shape accent judgments

Our ears deceive us in remarkably consistent ways, tilting our self-perception toward a version of ourselves that simply doesn't exist.

By: Graeme Brown

Jul 19, 2026 2 min read Editorial standards Meet our editors

Hearing your recorded voice can be unsettling because a recording removes part of the sound you normally hear when you speak.

Why the recording sounds different

Other people hear your voice mainly through air conduction. When you speak, you also hear vibrations conducted through the bones and tissues of your head. That additional route changes the mix of frequencies reaching your inner ear, which is why a recording can sound thinner or higher than the voice you recognize as your own.

Research published in Royal Society Open Science found that adding bone-conduction filtering to recordings improved people’s ability to identify their own voices. The result supports a specific point: bone-conducted sound contributes to vocal self-recognition.

Accent judgments involve a different effect

Expectations can also shape how listeners categorize an accent, but that is not the same mechanism. In a well-known experiment, Nancy Niedzielski played Detroit-area participants a recording of another Detroit speaker and asked them to match vowel sounds. Participants who were told the speaker was Canadian made different matches from those told the speaker was from Michigan, even though they heard the same recording.

That study showed that a social label can influence perception of another speaker’s vowels. It did not test how accurately people hear their own accents, and it does not show that self-perception of accent is permanently distorted.

A useful editing analogy, not a scientific equivalence

There is a practical analogy for writers. Authors bring intentions and background knowledge to their own sentences, while a new reader sees only what reached the page. An outside editor can therefore notice missing context that the writer supplied automatically.

The analogy should stay an analogy. Bone conduction is a physical feature of hearing; expectation effects in accent perception are social and cognitive; familiarity with a draft is a writing problem. All three remind us that self-perception and outside perception can differ, but the studies do not establish one universal mechanism connecting them.

Written by

Graeme Brown

Graeme Brown is Chief Editor and co-founder of Global English Editing, which he founded in 2014 with his son Brendan. He spent his career in education in Australia — first as a high-school mathematics teacher, then as an assistant principal — before moving into entrepreneurship and the editing world in his later years. He shares oversight of the editing team and the editorial standards on this site, and writes essays drawing on a long Australian life and a lifelong reader's instinct for the craft of sentences. His editorial focus is the patterns and habits of writing that hold up over a long career — what older writers know about finishing, the slow craft of revision, and what reading widely over decades quietly teaches you about how language actually works. He brings a teacher's instinct for what makes ideas land with readers and an editor's eye for the gap between what a writer means and what arrives on the page. Graeme is married to Jeanette Brown, who writes on the second half of life at jeanettebrown.net. They live between Cairns and Melbourne.

Why your recorded voice sounds unfamiliar, and how expectations shape accent judgments

Why the recording sounds different

Accent judgments involve a different effect

A useful editing analogy, not a scientific equivalence

Graeme Brown

More from Language & Intelligence

Real confidence rarely needs to announce itself: it shows up as stillness under pressure, a level voice when everyone else in the room is raising theirs, a willingness to let silence sit for a few extra seconds, and no need to perform for an audience that was never really watching

Received Pronunciation, the accent linguists still nickname the Queen's English, took shape inside a handful of England's most expensive private schools, and it remains such a reliable marker of that background that a stranger can often place someone's education within a single sentence.

University of Arizona researchers wired 79 people with a recorder sampling conversations every 12.5 minutes for four days and found the happiest had twice as many substantive conversations and a third as much small talk as the unhappiest, suggesting the discomfort was never a social deficit.

Psychologists did not coin emotional intelligence until 1990, and Goleman's book did not popularize it until 1995, so people raised between 1945 and 1975 grew up with no shared vocabulary for feelings, many still improvising one from watching, guessing, and going quiet when intimacy calls for words.

Why the recording sounds different

Accent judgments involve a different effect

A useful editing analogy, not a scientific equivalence

Graeme Brown

More from Language & Intelligence

Real confidence rarely needs to announce itself: it shows up as stillness under pressure, a level voice when everyone else in the room is raising theirs, a willingness to let silence sit for a few extra seconds, and no need to perform for an audience that was never really watching

Received Pronunciation, the accent linguists still nickname the Queen's English, took shape inside a handful of England's most expensive private schools, and it remains such a reliable marker of that background that a stranger can often place someone's education within a single sentence.

University of Arizona researchers wired 79 people with a recorder sampling conversations every 12.5 minutes for four days and found the happiest had twice as many substantive conversations and a third as much small talk as the unhappiest, suggesting the discomfort was never a social deficit.

A letter from our editors