Public Speaking
🗒️ Description
Hub for everything in this vault about giving talks: the researched playbook, the canonical books, and the toolchain. Three sources approach the same craft from different directions and disagree in useful places, so this note holds the map rather than the material.
I speak occasionally as a practitioner — conference talks of 25–45 minutes, community meetups, webinars, client sessions and cohort workshops, delivered in Polish to mostly technical audiences. The material here is cut for that, not for professional speaking.
🔗 Links
| Note | What it covers |
|---|---|
| Conference Talk Craft | The researched playbook: 27 techniques ordered by when they happen, with the evidence behind each one graded. Start here for anything operational. |
| TED Talks | Chris Anderson. The idea is the payload; trust before content; the five-step explanation ladder. |
| The Art of Public Speaking | Dale Carnegie, 1915. Emphasis and monotony, gesture from real feeling, the voice as a trained instrument, and an eight-question argument test. |
| Marp · Marp CLI · Claude Code Marp Workflow | The markdown-to-deck toolchain I actually use. |
🧩 Where the three sources agree
- Delivery is trained, not innate. Carnegie’s swimming metaphor and Anderson’s example of a shy 12-year-old coached into a strong TED talk in six months make the same claim a century apart.
- The fear is real and is not the obstacle. Both treat stage fright as something to master rather than eliminate. Gladstone and Beecher never lost it.
- Monotony kills a talk faster than weak content. Carnegie’s one-keyed piano; Anderson’s closing advice to vary the voice.
- One idea per visual, and only where the visual reveals something words cannot. Anderson states it directly; the research adds the cognitive-load evidence behind it.
⚔️ Where they contradict each other
Script or improvise. Carnegie says memorise at least the opening sentences. Anderson says a memorised talk sounds memorised, and reports that TED dropped its own no-scripts rule when Daniel Kahneman could not work under it. Conference Talk Craft resolves this the way I actually work: verbatim on the seams, free in the middle — which removes the single-point-of-failure risk on both sides.
Feeling versus structure. Carnegie’s load-bearing claim is that gesture and persuasion must emanate from genuine feeling, and that technique without it looks like branches nailed to a tree. The research is largely structural — checkpoints, cut lists, protocols. These are complementary, but the structural material can quietly crowd out the thing Carnegie says actually moves a room.
Arranging the audience. Carnegie recommends seating people densely so reaction spreads by contagion. At a conference the seating is not mine to arrange, and the framing is closer to crowd manipulation than I want to work in. Noted, not adopted.
⚠️ Read the evidence grading before quoting any of this
Roughly a third of the standard advice in this field is a sound practice attached to a debunked justification — the ten-minute attention rule, Mehrabian’s 7-38-55 ratio, power posing, the recency effect as a reason to close after Q&A. Both books above repeat at least one of these in passing. Conference Talk Craft carries the full list with sources, and the practices mostly survive once the wrong reason is removed.
📖 Further reading
- Research workspace with the full field-by-field data:
_raw/research-workspaces/wystapienia-publiczne-prelekcje/ - The Leader In You — the adjacent skill: the room after the talk rather than the stage
- Progressive Disclosure — the same layered-reveal principle, applied to knowledge bases
Template: knowledge_note_info