Topic 13
FreeHow You Sound
People take something from your voice before the sentence is over, and nervous speech mostly announces the nervousness rather than the content. These ten are mechanical: they change what your voice does rather than what you think. This is also the thinnest-evidenced topic in the academy, and it says so exercise by exercise — voice research is largely about persuasion and hiring rather than about being liked in a conversation.
This topic
10 to try
Next up — Take the breath first
The principle
Slow is not a style. It is the sound of somebody who expects to be listened to.
Evidence in this topic
2 Research2 Clinical practice6 Our readingTake the breath first
StretchBefore you speak in any group, take one breath low in your body, then start.
Why we teach it — Anxious speech starts on an empty tank, which is what makes it fast, high and thin. The fix sits upstream of the voice.
In the cognitive model of social anxiety, physical symptoms and self-monitoring feed each other: attention turns inward onto how you are coming across, which degrades how you come across.
What it cannot show — A theoretical model rather than an experiment. It organises what clinicians observed and has shaped the treatment of social anxiety ever since, but the individual mechanisms in it are not each tested by this paper — and it was written about diagnosed social phobia, while this academy applies it to ordinary shyness.
Land the ends of your sentences
StretchFinish sentences downward rather than upward. Say one out loud now, both ways, and hear it.
Why we teach it — A rising end asks for permission to have spoken. Most people do it hardest on their own name.
Our reading. Intonation research exists, but we found nothing clean enough to cite for this instruction, and we would rather say so than dress it up.
Pause instead of filling
StretchWhen you reach for “um”, close your mouth for the length of it instead.
Why we teach it — A pause reads as thinking. A filler reads as fear of losing the floor — and you do not lose it.
Our reading. Craft, from people who speak for a living.
But do not let the pause become a gap
StretchPause inside your own sentences, not between their turn and yours.
Why we teach it — The two are opposite things. A pause mid-sentence is composure; a gap before answering reads as distance.
Across conversations between strangers and between friends, shorter gaps between turns tracked how connected people felt — and third-party listeners could hear the difference too.
What it cannot show — Spoken conversation between friends and between strangers, with the gaps between turns measured in fractions of a second and connection reported straight afterwards. It says nothing about written messages, or about any pause long enough to notice.
Here — The effect lives in milliseconds of turn-taking, not in whether you take a second to think. It says a long gap costs something; it does not say hurry.
Say it aloud rather than writing it
StretchWhen you have something considered to say to somebody, say it. Do not draft it.
Why we teach it — The voice carries the thinking. The identical words on a page lose it.
When evaluators heard a candidate deliver a pitch rather than reading the identical words, they rated them as more intelligent, more thoughtful and more likeable, and were more interested in hiring them.
Here — Job-market pitches judged by recruiters and by untrained raters. The mechanism — a voice conveys a mind — is what is being borrowed.
Match the volume, not the energy
StretchSpeak at the level the room is speaking at. Do not try to match somebody else's brightness.
Why we teach it — A volume mismatch is what makes a conversation feel like work. An energy mismatch is just two different people, and that is allowed.
Our reading.
Slow the first sentence
StretchSay the first sentence of any conversation at about three-quarters of your natural speed.
Why we teach it — The first sentence sets the pace of the whole exchange, and nerves make it the fastest one you will say all evening.
Our reading. We know of no clean finding on speech rate and being liked in conversation; the ones that exist are about persuasion, and they disagree with each other.
Record yourself once
StretchRecord two minutes of yourself talking about something you like. Listen once. Notice one thing. Do not listen again this week.
Why we teach it — One listen is information. Five is self-surveillance, which is the thing this academy exists to reduce.
Post-event processing — replaying a social episode afterwards in detail — maintained anxiety rather than resolving it, and predicted more negative self-appraisal.
What it cannot show — Self-report after real social episodes, in samples selected for social anxiety. It establishes that the replaying happens and tracks distress; it does not show that stopping it removes the distress.
Here — Studied on the replaying of real social events rather than on deliberate practice. The limit of one listen is ours; the reason for it is theirs.
Read aloud for five minutes
StretchRead anything out loud for five minutes a day for a week. Poetry is best; a news article will do.
Why we teach it — It is the only exercise here that builds the mechanism directly, and it works on breath, pace and sentence endings at once.
Our reading. Standard practice among actors and broadcasters; we have no controlled evidence that it transfers to conversation.
Stop performing the last sentence
StretchWhen you finish, stop. No small laugh, no “anyway”, no summary of what you just said.
Why we teach it — The tail is where the confidence leaks out, and it is the easiest thing on this list to remove.
Our reading.
Sources cited in this topic
- 01Clark & Wells (1995). A Cognitive Model of Social Phobia. In Heimberg et al. (eds), Social Phobia: Diagnosis, Assessment and Treatment.
- 02Templeton, Chang, Reynolds, Cone LeBeaumont & Wheatley (2022). Fast Response Times Signal Social Connection in Conversation. Proceedings of the National Academy of Sciences, 119(4).
- 03Schroeder & Epley (2015). The Sound of Intellect: Speech Reveals a Thoughtful Mind, Increasing a Job Candidate's Appeal. Psychological Science, 26(6).
- 04Rachman, Grüter-Andrew & Shafran (2000). Post-Event Processing in Social Anxiety. Behaviour Research and Therapy, 38(6).
Where an exercise is labelled our reading, there is no study behind it and we have said so rather than borrowing authority from an unrelated one.
Where this leads
Playing