Public speaking, oriented

Every talk has parts you can plan for. You turn a thought into words and cues, the message travels by some medium, and each listener decodes it. Noise, meaning anything that blurs the message, can enter at either end, and the fix depends on its kind. In a large room, use a microphone instead of talking over side conversations. If listeners may read a word differently, swap the jargon for a plain one.

Hearing happens on its own; listening is work, and a phone or a busy setting easily breaks it. A reader can slow down and go back over a hard passage. A listener gets your words once, so repeat on purpose: say what is coming, develop it, then restate it at the close.

Meanwhile the audience keeps replying. Yawning, slouching and fidgeting may mean boredom; frowning, head-shaking or leaning back, disagreement or confusion; leaning in and nodding, engagement. Cues differ across cultures, so read them with care.

The trap is treating your plan as a script to finish no matter what. Say you are walking a group through a new procedure and see puzzled faces at step three. Pushing on to step four risks losing them. Slow down, check whether step three made sense, and ask an open question so listeners can say what is unclear.

Rule: plan for a listener who hears you once, and keep watching for their answers while you speak.

Communication apprehension is anxiety about speaking with others, felt during a talk or only in anticipation; it describes a feeling, not a level of skill. The pounding heart, dry mouth and clammy palms are the fight-or-flight response, and this anxiety tends to peak just before you begin and fade once the moment passes. One worry behind it is the illusion of transparency, the sense that your nerves are plain to everyone watching.

Avoiding the situation may bring brief relief, but untreated anxiety tends to stay. Graded exposure, the long-established alternative, starts with easier versions, such as a small, friendly group, and builds up. You can also notice a thought like 'everyone will see me freeze', challenge it, and put a more useful thought in its place. Reading a racing heart as energy is a nudge at best: a 2024 meta-analysis found a small gain that shrank by over a third once corrected for unpublished weak results.

Standard advice for nerves begins with getting ready: learn your facts and your audience, anticipate questions, and work out your first few sentences fully. Then stand and speak the talk from your outline several times, refining wording, emphasis and timing. Spread the runs over days; classroom studies pooled in a 2025 meta-analysis showed a moderate edge for spacing, though the effect is not universal. Time each run, use the real equipment, and record one to catch habits you cannot notice while speaking.

The trap is mistaking rereading for rehearsal. Silently going over notes the night before is not rehearsal; three timed, spoken run-throughs on different days are. On the day, glance over your plan, make no last-minute changes, and know your opening thoroughly. Afterward, if the talk will be given again, look back over it promptly, and distrust your own feeling about how it went.

Rule: treat nerves as normal and preparation as your first way to manage them, and rehearse aloud, standing, with the clock running.

Aristotle held that of speaker, subject and listener, the listener sets a speech's end, so analysis comes before content. Age, work and background are first guesses about a room, never conclusions; a federal training course on communication warns never to assume a diverse audience is all alike. Attitudes, beliefs and values shape how people hear you, and missing them can shut listeners out. The event counts too: room size, audience size, equipment, the time limit, and who speaks before and after you.

Where listeners stand changes your first goal. With a hostile room, older manuals advise earning a fair hearing before pressing the case: make people well disposed, find common ground and clear away prejudice. When you want listeners to act, show what they stand to gain; one federal course on communication advises naming one or two benefits instead of a long list.

Then fix the purpose. The general purpose is to inform or to persuade, sometimes to entertain, and a talk can mix them; the specific purpose says what this audience should know or do. A good speech advances one central idea and fits the clock: estimate how long each point needs and discard the rest, however good.

The trap is building from your own interests. Say you have ten minutes to tell a tenants' meeting about a new recycling program, and you love its history. They need to know which bin takes what and when collection starts. A purpose they can act on ends with a concrete step: what to do, how and when.

Rule: before you outline, write one sentence saying what these listeners should know or do when you finish, and cut whatever does not serve it.

Research converges when you know what you need. An early debate manual has you analyze the question first, because analysis reveals what you have to prove, and only then search for evidence on each main issue. Trace every fact to the most reliable source available: cite the official report, not the magazine that quoted it. Write down the full source as you take each note, never trusting memory to supply it.

Ask of every source who is behind it and whether they are qualified, whether it is reviewed and dated, whether it informs or sells, and who pays for it. Several items that trace back to one organization still lean on a single source, and enough evidence is not a count but whatever moves these listeners.

An example makes a claim concrete if it is typical, a comparison explains the new through the familiar if the likeness is real, and testimony borrows a witness's weight, most of all from someone with no biasing stake, full knowledge and expertise in that very field. Pick stories that represent the usual case, because a vivid anecdote can outweigh the data. What convinced you may not convince them, yet using evidence you know to be unreliable, just because they would accept it, is, in that manual's word, fraud.

Numbers are heard once. A federal guide on communicating risk reports whole numbers like 1 in 100 easier to grasp than decimals, comparisons easiest on a shared base, fewer numbers better, and relative changes more persuasive than absolute ones. Say a treatment takes a side effect from 6 cases in 10,000 down to 3. Calling that a 50 percent drop is true and tends to sway listeners more than the raw counts, so give the counts on their base too.

Rule: trace each key fact to its origin, choose support that will move these listeners, and give few numbers, whole and with their base.

Aristotle held that no one should use persuasion to make people believe what is wrong. The working standard has three parts: tell the truth, respect each listener's right to decide, and take responsibility for what your words do.

Plagiarism is presenting someone else's ideas, processes, results or words as your own. Since that covers ideas, rewording borrowed reasoning with synonyms or a new sentence order leaves it borrowed. A speech has no footnotes, so credit aloud: name who made a claim, not only where it was published, and say when the finding dates from.

Present evidence at the strength it has. Quote exactly and in context, since, as an early debate manual warns, well-chosen omissions can make a respected authority appear to back almost anything. Give the caveats a listener needs, such as how many people a study involved and who paid for it, and never pass off support as proof. If you hold a stake in what you urge, whether payment, a business connection or a personal relationship, tell the audience. Federal guidelines on advertising endorsements require disclosing such a tie when it could materially affect how much weight listeners give an endorsement and they would not expect it.

Where persuasion ends and manipulation begins is disputed. One influential view calls manipulation hidden influence, working in ways the audience is unaware of and could not easily notice; critics answer that open pressure, such as a guilt trip, can manipulate as well. Reasons a listener can weigh and reject leave the decision with them.

Say you are urging people to buy a product, and a company-funded study of a few dozen volunteers found a benefit. 'Research proves it works' inflates the evidence and hides the funder. Say who ran the study, how many took part and who paid, and mention your commission if you earn one.

Rule: give reasons people can weigh and reject, and, by the hidden-influence test, distrust any tactic that would stop working once explained to the audience.

A pattern is the path listeners follow. Chronological order suits histories, processes and instructions; spatial order moves through a place in one steady direction; topical order splits a subject into kinds or parts, still arranged for a reason. A causal pattern commits you to defending the causal link. Problem-solution proves the problem, where it came from and what it affects, before offering a remedy and showing that it is practical. Monroe's motivated sequence runs attention, need, satisfaction, visualization and action, so its close is built in.

Let purpose, audience and material pick, not habit: a grant-pitch guide warns that carrying a paper's methods-then-results order into a short pitch fails.

A short talk carries only a few main points; a 2025 guide to case-conference talks recommends the two or three most relevant. Phrase equal points in the same grammatical form. Give each part time by its weight, not in equal shares, and keep points at one level equally important, so an example sits under the point it illustrates. Listeners have no headings to look at, so signpost aloud: 'the second reason', 'turning now to cost'.

Build from a full preparation outline written in sentences, so you can see exactly what you claim, with one idea per entry and never a lone subpoint. Then speak from a short outline of key words and any quotation you must give exactly, since fuller notes drift back into reading.

Say your outline reads: I. Background, II. Features and benefits, III. The March outage. Point II holds two ideas, and III is an example posing as a main point. Split II into what the system does and what it saves, and move the outage under the point it proves.

Rule: if the outline makes no sense stripped of its prose, fix the skeleton before you draft.

Attention is highest at the start, Aristotle observed, so spend the opening on substance. A story, a question, a startling claim or a quotation can open a talk. An early-1900s manual warns, though, that a surprise used only as a trick may catch attention for a moment and still fail to keep a thoughtful audience. Then tie the topic to your listeners' own stake, since people attend to what concerns them, and say briefly what connects you to it, keeping only the experience that shows expertise. State your thesis, the single sentence you would save if the rest had to go, and preview your main points in order, promising only what you will deliver.

Two openings often misfire. An apology is dull, and admitting you are unprepared spends your credibility before you start. A joke dragged in for its own sake can miss both topic and occasion.

A conclusion lands the talk. Restate your central point in slightly different words, so it sounds like an arrival rather than a repeat, and recap the main points, since listeners cannot reread. If you want action, name a concrete act; a 19th-century manual warns that an appeal so general it could close any speech helps none of them. A short final line, one early manual says, can cut deeper than a long one. Once you signal the end, end, adding nothing new after 'in conclusion'.

Transitions should say how two parts relate, not just announce a new one: 'another point' merely announces, while 'but' and 'so' tell listeners how to take what follows.

Say you open with, 'Sorry, I only heard yesterday that I was speaking.' Replace it with why the topic matters to this room, then your thesis and a preview.

Rule: open with why it matters and what you claim, close by restating it in new words, and after one clear signal, stop.

Speech and print reward different styles. Aristotle noticed that repetition and runs of clauses without connecting words, faulted on the page, work in live delivery. So write to be heard: everyday words, contractions, short sentences and repeated key words. Concrete, picture-making words matter even more aloud than in writing. Vary sentence length, since a run of short ones turns monotonous and a long one loses the hearer unless pauses break it, and give your key line a short sentence.

In Aristotle's account, a metaphor lets a listener grasp something new through a trait it shares with something known, provided it is neither far-fetched nor obvious. Metaphors shape reasoning too: in one set of experiments, one metaphor word describing crime shifted the solutions readers preferred, and few noticed the effect. Keep the technical terms you truly need and define each where you use it. On disability, prefer people-first wording, which names the person before the condition, and ask people how they want to be described.

Aristotle treated style as an art of its own: knowing what to say is not enough; you must say it as you ought. Clarity comes first, from ordinary current words, with a few unusual ones to keep speech from going flat. Style should fit the subject and the speaker, since dressing a plain noun in ornate epithets turns comic. He called strained compounds, obscure words, piled-up epithets and far-fetched metaphors frigid.

The trap is drafting for the page. Say your draft reads, 'Implementation of the initiative is anticipated to yield significant efficiencies.' Aloud, try: 'This plan saves time. A lot of time.'

Rule: write for the ear, with plain words, varied sentence lengths, and figures of speech only where they fit and clarify.

Reading a manuscript tends to sound dull and cuts eye contact, and a memorized talk often sounds recited. Extemporaneous speaking is planned and outlined, then worded fresh as you speak; an Air Force guide uses it for most briefings and keeps manuscripts for occasions when the exact words matter.

Your voice is a set of controls. Volume comes from breath support, not a strained throat, and a crowd absorbs sound, so check with someone in the back row that you can be heard. Vary your pace, faster for excitement and slower for weight, and if you rush, rehearse at a speed that feels too slow. A pause beside a key word gives it weight and replaces the 'uh' that fills a gap, and it feels longer to you than to listeners. End claims you stand behind on a falling pitch, since a rising one suggests doubt.

Look at people in every part of the room, not at notes or the screen. Use gestures that picture or point at your meaning; repeated, aimless ones such as hand-wringing draw attention to the habit. Stand balanced, move to a new position as you start a new point instead of pacing nonstop, and ask an honest observer about rocking or leaning you cannot see yourself.

A slide earns its place by showing what speech cannot, such as a timeline or a comparison. Give each one message, ideally as its heading, and keep text sparse; studies of slide text and recall are mixed, but short phrases that echo you are a safe middle. Never rely on color alone, and describe aloud what is shown.

Say your slide holds six full sentences and you turn to read them. You have made yourself redundant. Cut it to one heading, face the room, give people a moment to read, then talk.

Rule: use notes and slides in a way that keeps your eyes on the audience.

Explaining shows what a subject is and how it works; arguing trades in reasons and inferences toward a verdict. A 1920 textbook notes that a clear explanation can sometimes win agreement on its own, so check each section of your draft: once it urges a verdict, it has crossed into argument.

Let the subject pick the order: time for processes and history, place for layouts, division for a topic made of kinds. Start from what listeners know and assemble the new idea one plain fact at a time, with typical examples and comparisons whose resemblance is real. And cut: saying too much buries the main message, speeds you up and shrinks question time.

In persuasion, whoever proposes changing things as they are generally carries the burden of proof. Listeners with the motivation and ability to think hard about your case weigh its arguments, while others lean on cues such as who is speaking, and research summaries report that change built on argument tends to last.

Controlling wording like 'you must' can provoke pushback: in one experiment with 1,350 adults, strongly threatening wording lowered intentions when paired with talk of losses but raised them when paired with gains. Whether fear appeals work is debated, but researchers broadly agree that showing a doable way to deal with the threat helps. Telling listeners that few people do the right thing can backfire by making the wrong thing look normal; a 2024 review suggests stressing instead that the good behavior is spreading. In a reanalysis of 107 studies, mentioning the other side made about zero difference on average, with wide swings, so do not count on the format, and if you raise an objection, answer it.

Say you want a team to start backing up its files. 'Hardly anyone here backs up, and you must start' stacks a norm that can backfire on top of a command. Try: 'More of us are setting up backups, and it takes two minutes; here is how.'

Rule: know whether each section explains or argues, and when you argue, give reasons people can weigh and one step they can take.

Five shapes of argument are worth knowing. Induction builds a general claim from cases and fails by hasty generalization, so look for the case that breaks your rule. Deduction applies a general rule to a case; true, well-linked premises carry the conclusion, so critics usually attack the premises, a 1920 textbook notes. Aristotle noted that speakers often skip a premise the audience already knows. A causal claim needs more than sequence: the cause may run the other way, or a third factor may produce both. An analogy holds only where the two cases agree on the points that bear on the conclusion. A sign, like a packed parking lot suggesting a busy store, usually has rival explanations. Many arguments also lean on an unstated link from evidence to claim, which informal logicians call a warrant.

Aristotle defined rhetoric as the ability to see the available means of persuasion in each case: finding everything persuasive the case allows, not guaranteeing a win. On that view, judge your preparation by whether you left a usable argument unused, not only by the result. He named three appeals. Logos is the argument itself; a crowd cannot follow a long chain, so start from what they already believe and keep the steps few. Ethos is your character as the speech reveals it, earned in what you say rather than borrowed from reputation, and resting on good sense, good moral character and goodwill. Pathos is the listener's feeling, which matters because the same facts look different to friendly and hostile listeners; fear, for instance, needs a danger that feels near and likely, with some hope of escape. Yet he condemned warping a judge with anger or pity.

Say you claim that sales rose after the new website launched, so the website worked. Your warrant is that whatever came first caused the rise. Said aloud, it fails until you rule out other causes.

Rule: say your hidden premise aloud, ask what else explains the evidence, and let any feeling you raise come from the facts at issue.

Older manuals single out one failing across special occasions: length. First ask how long you have and who speaks before and after you. Introducing a speaker means readying the audience, showing why this speaker fits and stepping aside, with every name and title checked. Presenting an honor names why it is given; accepting is mostly thanks, kept short. A toast should be brief and aimed at the person honored. A eulogy shows a few real deeds that reveal character, in plain words; a 1920 textbook warns that grandeur borrowed from famous eulogies sounds like bombast. Commemorative speeches heighten facts the audience already accepts rather than proving new ones.

On the spot, you still need a plan. A federal crisis-communication manual tells spokespersons to lead with the key point, in roughly 30 seconds and fewer than 90 words, back it with prepared support, then stop. Hear the whole question before composing a reply, and split a multipart one. Let an angry questioner vent for a reasonable time, acknowledge the feeling, then address the issue rather than anyone's character. Bridge to your key message only after you have answered. If you do not know, say so, commit to finding out, and follow up as promised.

On a panel, settle in advance who takes which topics. If time runs short, drop a section you chose in advance rather than speeding up.

Say someone asks, 'Why did your team waste the budget on this?' Do not repeat 'waste', even to deny it. Restate it neutrally, 'You are asking how the budget was spent', and answer that.

Rule: answer first, briefly and honestly, and stay inside the time you were given.