Public Speaking

Audience

Learners of any age or experience level who want low-pressure ways to practice speaking, listening, recovery, and revision—with or without AI.

Purpose

Give learners repeatable ways to think, listen, speak, recover, and revise without turning delivery into a personality judgment or a single score.

Speaking practice makes evidence boundaries visible: feedback can sound convincing even when a person or tool did not observe the performance. Learners practice “trust, but check” by asking what the coach actually received before deciding what to use.

Learning Outcomes

Learners can:

• Set clear conditions for a speaking attempt.
• Make an attempt without treating fluency as proof of understanding.
• Reflect before receiving feedback.
• Ask what evidence a coach observed.
• Choose one change for the next attempt.
• Make a second attempt and compare it with the first.
• Separate delivery style from personality, identity, or ability.

Shared Practice Loop

Set conditions → attempt → learner reflects first → specific soft observations → learner chooses one change → second attempt → compare.

The learner owns the change. A coach, partner, observer, transcript tool, or AI system may contribute observations only within what it actually received.

Evidence-Bounded Coaching

Every coach must disclose whether it observed audio, transcript, or human notes.

• A plain-text transcript cannot assess pauses or pace. A timestamped transcript may support estimates of pace or pause length, but a transcript cannot establish volume, pronunciation, or expression.
• A listener may describe named cues—such as an audible pause, a repeated phrase, a clear example, or a difficult transition—without claiming to read personality or emotion.
• “Perceived confidence” is permitted only as a soft listener impression tied to named delivery cues. It is not a fact about the speaker.
• Do not assign grades, a single score, personality or emotion claims, accent grading, or unsupported winner declarations.
• Feedback begins with what the learner noticed and ends with one change the learner chooses.

For example:

• Useful: “I heard a pause before the example, which helped me follow the transition.”
• Unsupported: “You sounded nervous and lacked confidence.”

Mixed-Age Controls

Adjust vocabulary, preparation time, speaking duration, interruption level, twist intensity, humor, and non-speaking roles. Change one dimension at a time. Do not infer confidence, ability, topic, humor, or technology comfort from age or generation.

Choose a First Practice

Start with the first practice unless another option better fits the learner’s goal.

• Forty-Five Seconds to Make It Clear (legacy Learning Lab) — recommended first. Turn one idea into a short explanation for a real audience, compare audience-specific versions, and explain the choices. Shorter and extended routes are also available. This remains primarily a Conversation Engineering activity; this module uses it as an accessible first speaking practice.
• One Minute to Think, Forty-Five Seconds to Speak (legacy Learning Lab) — Flexible. Prepare, speak twice, reflect first, choose one change, and compare the attempts.
• Reading Lines Together (legacy Learning Lab) — Flexible. Mark cues in an RS-authored scene, read it twice, and revise one delivery choice without grading accent or personality.
• Improv: Accept, Add, Continue (legacy Learning Lab) — Flexible. Practice listening, accepting a premise, recovering, and choosing one change across two scene attempts.
• Debate Practice (legacy Learning Lab) — Flexible. Build claims, responses, and closings across two rounds while keeping observations tied to what listeners heard.

Relationship to the Tracks

Public speaking supports Invitation To Be Curious Again by making room for experiments, recovery, and learner-chosen revision. It supports AI Literacy for the Rest of Us by turning “trust, but check” into a practical habit: before using human or AI feedback, learners ask what the coach observed and what its evidence can support.

Rights and Respect

Use RS-authored scenes, learner-authored text, public-domain material, or text you have permission to use. Follow its license or permission terms, keep private material private, and do not republish it through the Learning Lab. Do not imitate a recognizable living or deceased person’s voice. Accent and dialect are not defects to grade.

Tool Boundary

AI is optional in every practice. A human partner, observer, printed script, text rehearsal, facilitator demonstration, or non-speaking role may meet the same outcome. Before accepting tool feedback, identify whether the tool received plain text, timestamped text, audio, or human notes, and limit its claims to that evidence.

Privacy Boundary

No recording is required. Any optional recording remains learner-controlled and outside the Learning Lab. Get clear permission before recording another person or sharing their speech or transcript with a tool. Upload only material you own or have permission to use, and choose the lowest-risk input that supports the practice.