Organizing and Outlining
You have a topic, a purpose, a thesis, and a folder of good sources. None of that is a speech yet. A speech is a sequence — one idea after another, arriving in an order your listeners can follow the first time, in real time, with no way to back up. Organizing is the work of turning a pile into a path. Outlining is the work of proving to yourself that the path holds together before you stand up and walk it. This chapter is unapologetically procedural. You will learn how to pull two to four main points out of a thesis, how to pick a pattern that does some of your arguing for you, how to build the verbal handrail of transitions and signposts that keeps an audience oriented, how to write a full-sentence outline that survives its own logic tests, and how to boil that outline down to a speaking outline you can actually glance at while your heart is pounding.
By the end of this chapter you can
- Derive two to four main points directly from a thesis statement and defend why those points, in that order.
- Apply four tests — weight, overlap, sufficiency, and count — to diagnose a weak set of main points and repair it.
- Select and justify an organizational pattern (chronological, spatial, topical, causal, comparison, problem-solution) for a specific audience and purpose.
- Build a persuasive body using Monroe’s motivated sequence and identify which step speakers most often skip.
- Write transitions, signposts, internal previews, and internal summaries that name the seams of your speech out loud.
- Construct a full-sentence outline that obeys coordination, subordination, the rule of division, and parallelism.
- Convert a full-sentence outline into a speaking outline with keyword phrasing and embedded delivery cues.
- Time a draft and cut it by a third without gutting the argument, using modular subpoints.
Section 1From Pile to Path: Finding Your Main Points
The listener has no scroll bar
A reader who loses the thread can stop, back up half a page, and re-enter. A listener cannot. Your audience gets one pass, at your speed, in a room with a humming vent and someone’s phone buzzing. Every navigational job that a reader does with their eyes — skimming headings, checking paragraph breaks, flipping back to the definition — a listener has to be handed by you, out loud, in the structure of the talk itself. That is the entire argument for organizing. It is not tidiness. It is accessibility.
Psychologists have argued for seventy years about how much a person can hold in mind at once. George Miller’s famous 1956 paper proposed a limit of about seven items, plus or minus two. Later researchers, notably Nelson Cowan, argued that when you strip away rehearsal and grouping tricks, the real working-memory limit is closer to four. The exact number is still genuinely contested, and honest instructors should say so. But every version of the finding points the same practical direction: fewer, bigger, clearly labeled chunks beat many small ones. For a five- to eight-minute speech, that means two to four main points. Not six. Not “I have seven things and they are all important.”
There is also older experimental work — studies from the 1960s and 1970s by researchers such as Thompson and by Sharp and McClung — suggesting that scrambling a speech’s structure reduced both comprehension and listeners’ ratings of the speaker’s competence. Those studies are decades old, the effects were not enormous, and later work has complicated them. Treat the claim modestly: disorganization probably costs you some comprehension and some credibility. That is reason enough to fix it, and you do not need a dramatic effect size to justify twenty extra minutes of planning.
Harvesting main points from the thesis
Chapter 6 asked you to write a thesis. Now make it work for a living. A well-built thesis already contains the shape of the body; your job is to read it back and notice what it promised. Take this one, written by a working nurse for a five-minute informative speech:
“Sepsis kills faster than almost anything else that walks onto a medical-surgical floor, and any bedside nurse can catch it early by knowing three cues, escalating without waiting for permission, and documenting exactly what was seen.”
Read that sentence with a pen. It promises cues, it promises escalation, and it promises documentation. That is three main points, already coordinate, already in a defensible order — you have to notice something before you can escalate it, and you escalate before you chart it. The thesis did the outlining. This is the payoff for writing a specific thesis instead of a vague one: a thesis like “sepsis is a serious problem” promises nothing and therefore organizes nothing.
When a thesis does not hand you the points that cleanly, generate options rather than settling for the first idea. Write the same topic three ways and choose deliberately. The table below shows one thesis and three legitimate skeletons; none is wrong, but each asks the audience to do something different.
| Skeleton | Main points | What it asks the audience to do |
|---|---|---|
| Topical (by cue) | I. Temperature and heart rate together. II. Mental-status change. III. Urine output falling off. | Memorize a short checklist they can run at the bedside. |
| Chronological (by hour) | I. Hour zero: the subtle presentation. II. Hour two: the missed window. III. Hour six: what the numbers look like when it is late. | Feel the cost of delay as a story unfolding in time. |
| Problem-solution | I. Early sepsis is routinely missed on busy floors. II. The reason is that the cues look like ten harmless things. III. A thirty-second screen at every vitals check closes the gap. | Adopt a specific behavior change on their next shift. |
Four tests for a set of main points
Once you have candidate points, audit them before you write a word of body text. Four quick tests catch most structural problems.
The weight test (coordination). Are your points roughly equal in scope and importance? Points at the same level of an outline should be siblings, not a parent and a child. “Causes of nurse burnout” and “One hospital’s twelve-week mindfulness pilot” are not siblings; the second belongs under a solutions point, not beside it. The overlap test. Can you say each point without wandering into the next? If your second point keeps forcing you to explain your third, they are one point wearing two hats. The sufficiency test. If a listener believed all your points and nothing else, would the thesis be delivered? If something essential is missing, you have a hole; if a point is not needed to deliver the thesis, you have a passenger. The count test. Two to four for a short speech, three for most of them; go past four only when the speech is long and the points are genuinely parallel and short.
Here is the audit in action. A student arrived with seven “main points” for a six-minute speech on hospital alarm fatigue. Watch what happens when the tests are applied: nothing is thrown away, but five of the seven get demoted into support, where they finally have room to breathe.
| Before — seven flat points | Diagnosis | After — three points with support |
|---|---|---|
| 1. Nurses hear 700 alarms per shift. 2. Most alarms are false. 3. Nurses stop responding. 4. A real alarm gets missed. 5. Patients die. 6. Better default settings help. 7. Tiered alarm systems help. | Fails the weight test (items 4 and 5 are consequences of item 3, not siblings) and the count test. Items 6 and 7 are two versions of one solution. | I. The volume is the problem (support: 700 alarms, roughly 85 to 99 percent non-actionable in published audits). II. Volume produces desensitization (support: the delayed-response pattern, one sentinel-event example). III. Two fixes are already available (support: patient-specific defaults; tiered escalation). |
A specific thesis hands you your main points; your job is to test them for equal weight, no overlap, full coverage, and a count a listener can actually hold.
Section 2Patterns That Do Some of the Arguing for You
The six workhorses
You do not have to invent a structure from scratch. Six patterns cover the overwhelming majority of classroom and workplace speeches, and each one carries a built-in logic that the audience already recognizes. When you choose a pattern, you are choosing what kind of thinking you want your listeners to do.
| Pattern | Organizing principle | Best for | Sample main points |
|---|---|---|---|
| Chronological | Time, or the steps of a process in order | How-to speeches, histories, timelines, protocols | How a trauma bay uses the first ten minutes: I. Minutes 0–2, primary survey. II. Minutes 2–5, adjuncts and imaging. III. Minutes 5–10, the disposition call. |
| Spatial | Physical arrangement, location, or depth | Anatomy, floor plans, geography, layered structures | Staging a pressure injury: I. Stage 1, intact skin. II. Stage 2, partial thickness. III. Stage 3, into fat. IV. Stage 4, to bone or tendon. |
| Topical | Natural categories of the subject | The default when nothing else fits; lists of types, barriers, benefits | Three barriers to hand-hygiene compliance: I. Time pressure. II. Skin breakdown. III. Dispenser placement. |
| Causal | Cause to effect, or effect back to cause | Explaining why something happens; setting up a solution later | Why rural hospitals close: I. The reimbursement squeeze (cause). II. Service-line collapse and longer transport times (effect). |
| Comparison-contrast | Two or more options held side by side | Recommendations, product or method choices, myth-versus-fact talks | Teaching insulin injection: I. The demonstration-only method. II. The teach-back method. III. Where they diverge on retention. |
| Problem-solution | Harm, then remedy (often with a cause step between) | Persuasive speeches asking for a policy or behavior change | Alarm fatigue: I. The problem on telemetry units. II. Why the defaults create it. III. What a tiered system would change. |
Matching the pattern to the audience, not to the topic
Students usually ask “which pattern goes with my topic?” That is the wrong question. Most topics accept several patterns. The right question is: what does this particular audience already believe, and what do I need them to do? Chapter 5 asked you to analyze the audience; this is where that analysis pays.
Take one topic — hospital nurse-to-patient staffing ratios — and three audiences. For a class of first-semester nursing students who have never worked a floor, chronological works: walk them through one nurse’s twelve-hour shift at a 1:4 ratio and then at 1:7, and the argument makes itself. For a room of experienced nurses who live the problem daily, problem-solution is a waste of their first ninety seconds; they know the problem. Jump to comparison-contrast: California’s mandated ratios versus committee-based staffing plans, held side by side. For hospital administrators who dispute the premise, causal is the honest choice: establish, with evidence, that the ratio causes the outcome, because unless they accept the causal link, no solution you propose will land.
Monroe’s motivated sequence, worked all the way through
Developed by Alan Monroe at Purdue in the 1930s, Monroe’s motivated sequence is a five-step persuasive pattern that follows the natural arc of a listener deciding to act. It is not problem-solution with two extra steps bolted on; the visualization and action steps are the whole reason it exists. Note that steps one and five overlap with what Chapter 9 calls the introduction and conclusion, so in a real outline the sequence spans the entire speech, not just the body.
| Step | Its job | Sample line — asking colleagues to attend post-code debriefs |
|---|---|---|
| 1. Attention | Earn the right to the next sixty seconds | “Last March, our unit ran a code at 3 a.m. By 3:40 the room was clean, the family had gone home, and six of us went straight back to our assignments. Nobody said a word about it. Ever.” |
| 2. Need | Establish the problem and make it theirs | “Unprocessed critical incidents are one of the documented drivers of turnover in acute care, and turnover on this unit ran twenty-two percent last year.” |
| 3. Satisfaction | Present the solution concretely enough to picture | “A structured debrief takes fifteen minutes, follows a five-question script, and is led by whoever was charge that night.” |
| 4. Visualization | Show the future with the solution, or without it | “Picture next March. Same code, same 3 a.m. But at 4:15 the six of you are in the break room, and one of you finally says the thing you have all been carrying.” |
| 5. Action | One specific, doable, immediate ask | “The sign-up sheet is on the break-room door. Put your name on one date this month. One.” |
The step students skip is visualization. It feels redundant — you just explained the solution, so why describe it again? Because explaining is not the same as picturing, and people act on pictures. The step students botch is action. “So think about it” is not an action step. “Learn more about this issue” is not an action step. An action step names a behavior, a time, and a place, and it is small enough that a tired person can say yes to it.
Pick your pattern from what the audience already believes and what you need them to do, not from the shape of the topic itself.
Section 3Connective Tissue: Saying the Seams Out Loud
Four connectives and what each one does
On the page, your structure is visible — indentation, numerals, white space. In the air, all of that vanishes. Connectives are the words that make an invisible outline audible. There are four, and they are not interchangeable.
A transition is a bridge sentence that closes one point and opens the next, usually by naming both. A signpost is a short verbal marker of position or emphasis — “second,” “the most important of these,” “here is the part that surprised me.” An internal preview tells the audience what is coming inside a section before you deliver it. An internal summary restates what a section just established before you move on. Previews reduce the load of listening; summaries convert listening into memory. Long or complex points deserve both.
| Weak version | Why it fails | Stronger version |
|---|---|---|
| “Okay. So. Anyway, moving on.” | Marks that something changed but not what or why. The audience files nothing. | “So the cues are subtle and easy to explain away. That is exactly why the next part matters: what you do in the ninety seconds after you notice them.” |
| “My next point is about documentation.” | Announces the speaker’s agenda instead of connecting ideas. Nothing links back. | “You have escalated. Now the only record that you saw it in time is the one you write — which brings us to charting.” |
| “There are a lot of things to say about burnout.” | A false preview: it previews quantity, not content. | “Burnout shows up in three places on this unit: in the schedule, in the charting, and in the way we talk to each other at shift change.” (internal preview) |
| “And that’s pretty much it for causes.” | Signals an ending without consolidating anything. | “So: mandatory overtime, a documentation burden that grew forty percent, and no debrief structure. Three causes, all of them local, all of them fixable.” (internal summary) |
| “First… second… also… and another thing…” | Numbered signposts that stop being numbers. The audience loses count and stops trying. | “First… second… and third, the one I want you to leave with…” Keep the numbering honest and finish it. |
Where connectives go
Place a transition at every seam: after the introduction into the first main point, between each pair of main points, and out of the last main point into the conclusion. In a three-point speech, that is four transitions, and they should be written down word for word in your outline. Transitions are the single most common thing speakers improvise badly under pressure, precisely because they are the moments when you are also thinking about what comes next. Write them out. Nothing else in the outline earns as much stability per sentence.
Add an internal preview at the top of any main point that has three or more subpoints, and an internal summary at the bottom of any point that carried numbers, a mechanism, or a definition the rest of the speech depends on. A useful rhythm for a dense point: preview, subpoint, subpoint, subpoint, summary, transition. It sounds repetitive when you read it silently. It sounds clear when it is spoken once at conversational speed to people who have never heard it before.
Repetition that helps and repetition that grates
New speakers fear repeating themselves. Experienced speakers repeat constantly — and the skill is knowing what to hold fixed. Hold the structural terms fixed: the point labels you promised in the preview come back word for word in the body and again in the summary, because that repetition is the thread an audience follows (Ch. 9 and Ch. 13). Vary the language around them. Restating a claim in the identical padded sentence feels like a stall; the same label attached to a new concrete detail feels like reinforcement. Compare: “So again, hand hygiene is important. Hand hygiene is really important.” against “So: twenty seconds, every time, even when the room looks clean. That is the whole ask.” The second keeps the ask fixed while advancing the language around it, and it converts an abstraction into a behavior.
One more distinction worth keeping straight. A signpost is not a filler. “Um, so, like, the next thing” is filler; “here is the part that changed how I chart” is a signpost. Fillers arise when your mouth outruns your plan. The best cure for filler is not willpower — it is a written transition, because you cannot stall in a place where you already know exactly what sentence comes next.
Write your transitions word for word and preview or summarize any point dense enough to lose someone, because structure the audience cannot hear does not exist.
Section 4The Full-Sentence Outline: Proving the Speech Before You Give It
Format and the four rules
The full-sentence outline — sometimes called the working or preparation outline — is the document where you do your thinking. Every main point, subpoint, and sub-subpoint is written as a complete sentence, sources are cited in place, and connectives are written out. It is long. It is supposed to be. It is not what you take to the lectern.
The standard alphanumeric format ranks ideas by symbol and indentation: Roman numerals (I, II, III) for main points, capital letters (A, B, C) for subpoints, Arabic numerals (1, 2, 3) for supporting material, lowercase letters (a, b, c) below that. Four rules govern the whole document.
Coordination: items at the same level have roughly the same importance. Subordination: each indented item supports the item above it — if you can put “because” or “for example” between a parent and its child and it makes sense, subordination is sound. Division (the rule of division): if you divide something, you get at least two parts, so an A requires a B and a 1 requires a 2. A lone subpoint means you have either restated the parent or hidden a second subpoint you have not written yet. Parallelism and singularity: siblings share grammatical shape, and each entry contains exactly one idea — if a line has an “and” in the middle joining two claims, it is two lines.
Diagnosing a broken outline
Below is a real-shaped excerpt from a first draft, with the repair beside it. Read the left column and try to name the violation before you read the middle.
| Faulty draft | Violation | Repaired |
|---|---|---|
| I. Nurse burnout is bad. II. Twelve-hour shifts. III. What our hospital could do about it and why the last attempt failed. | No parallelism (a claim, a noun phrase, a compound question), and III holds two ideas. I is too vague to support. | I. Burnout on this unit is measurable, not anecdotal. II. Three scheduling practices drive most of it. III. Two of those three can be changed at the unit level. |
| II. Three scheduling practices drive most of it. A. Mandatory overtime is the largest driver. | Faulty division — a lone A. You promised three practices and named one. | II. Three scheduling practices drive most of it. A. Mandatory overtime is the largest driver. B. Back-to-back rotating shifts prevent sleep recovery. C. Self-scheduling closes before part-timers can bid. |
| I. Burnout is measurable on this unit. A. Turnover ran twenty-two percent last year. B. Our hospital opened in 1974. | Faulty subordination — B does not support the main point above it. It is true and irrelevant. | I. Burnout is measurable on this unit. A. Turnover ran twenty-two percent last year. B. Exit interviews named workload in eleven of nineteen departures. |
| A. Overtime is bad and it also causes errors and people quit. | Not singular — three claims in one line, so none can be supported properly. | A. Mandatory overtime raises error rates. B. Mandatory overtime predicts intent to leave. |
What a finished excerpt looks like
A full-sentence outline holds more than main points. Put the oral citation right where you will say it, so you never have to remember mid-sentence where a number came from. Write connectives in brackets. Some instructors ask for a running time estimate in the margin; it is a habit worth keeping even when it is not required.
- II. Three scheduling practices drive most of the burnout on this unit. (about 90 seconds)
- [Internal preview] “Overtime, rotation, and who gets to pick first.”
- A. Mandatory overtime is the largest single driver.
- 1. In a 2016 review in the International Journal of Nursing Studies, Dall’Ora and colleagues reported evidence linking shifts of twelve hours or longer to burnout. [oral citation — say the source out loud, as a sentence]
- 2. On our unit, fourteen of twenty-two nurses worked at least one mandated shift last quarter.
- B. Rotating between days and nights within the same week prevents sleep recovery.
- 1. Sleep researchers describe circadian re-entrainment as taking roughly one day per hour of shift.
- 2. Our current template rotates within four days.
- C. Self-scheduling closes before part-time staff can bid, pushing them into leftover shifts.
- [Internal summary] “Overtime, rotation, and bidding order.”
- [Transition] “Those are the three. Here is the good news: two of them are ours to change without a single new hire.”
Notice what this excerpt makes impossible. You cannot hide a missing citation. You cannot drift out of parallel. You cannot discover, thirty seconds before you speak, that point B has no evidence under it. That is the entire value proposition of the full-sentence outline: it fails on paper, cheaply, in private, instead of failing out loud.
A full-sentence outline exists to expose weak logic and missing evidence while they are still cheap to fix, which is why every line is a complete sentence and every lone subpoint is a warning.
Section 5The Speaking Outline and the Rehearsal Loop
Boiling sentences down to cues
The speaking outline (also called a keyword or delivery outline) is what you actually hold. It exists to answer one question fast: what comes next? Full sentences defeat that purpose, because a page of sentences invites you to read, and reading kills eye contact, vocal variety, and the sense that you are talking to human beings. Keywords force you to look up and speak.
Two things stay in full form no matter what: direct quotations and your first and last sentences. Statistics are the middle case — they compress to the number plus enough of the source to say it accurately, never to the number alone. Everything else compresses freely.
| Full-sentence outline line | Speaking-outline line | Delivery cue in the margin |
|---|---|---|
| Mandatory overtime is the largest single driver of exhaustion on this unit. | OT = biggest driver | slow down |
| In a 2016 review in the International Journal of Nursing Studies, Dall’Ora and colleagues linked shifts of twelve hours or longer to burnout. | Dall’Ora 2016 review, Intl Jrnl Nursing Studies — 12-hr shifts → burnout | say the source aloud |
| Fourteen of our twenty-two nurses worked at least one mandated shift last quarter. | 14 of 22 — last quarter | pause after |
| “Those are the three. Here is the good news: two of them are ours to change without a single new hire.” | [TRANSITION — verbatim] Those are the three… two are ours to change. | step left, then speak |
| Put your name on one date this month. One. | SIGN-UP SHEET. One date. ONE. | stop moving; eye contact; end |
The physical page
Design the page for a person whose hands are shaking slightly and whose eyes have three-quarters of a second to spare. Print in a size larger than feels reasonable — eighteen to twenty-four point. Use one side of the paper or card only; flipping is visible and it costs you a beat. Number every card in the top corner, because dropped cards are not a hypothetical. Leave generous white space; a dense page has no landmarks. Put delivery cues in a different color or in italics so you never mistake a stage direction for something to say out loud — and yes, people do read “pause here” aloud when they are nervous.
Cards or paper is partly personal preference and instructors genuinely disagree. Cards are discreet and easy to hold; a single sheet on a lectern cannot be shuffled out of order. What both camps agree on: never bring the full-sentence outline. It contains the temptation to read, and under pressure you will take it.
Time it, then cut it
Almost every first draft runs long. Time your rehearsal with a real clock, not an estimate. If you are over, resist the temptation to speak faster; you will only sound rushed and lose the pauses that make structure audible. Cut instead, and cut structurally.
Modular outlining makes cutting nearly painless. Because your support sits in discrete subpoints, you can remove the second example under point I, the third statistic under point II, and one of two anecdotes under point III, and the architecture never moves. What you must not do is delete a main point at the last minute — that breaks your preview, one transition, and your conclusion in one stroke, and those are the pieces you know least well. Trim the leaves, not the branches.
Use this checklist for the final pass before you walk in.
| Check | What passing looks like |
|---|---|
| Point count | Two to four main points; you can say all of them from memory in ten seconds. |
| Pattern | You can name your pattern and say in one sentence why it fits this audience. |
| Division | No lone A, no lone 1, anywhere in the outline. |
| Connectives | Every seam has a written transition; every dense point has a preview or summary. |
| Citations | Each source is written where it will be spoken, naming the source, its authority and the date before the claim (Ch. 7). |
| Timing | Rehearsed aloud, standing, at or under the limit with roughly ten percent of margin. |
| The page | Keywords only, one side, numbered, cues visually distinct from spoken text. |
Convert the working outline into a keyword page you rehearse from out loud, and when you run long, cut support rather than structure.
The chapter in one breath
Organizing is the courtesy you extend to people who get exactly one pass at your ideas. It begins with a thesis specific enough to hand you its own main points, and it survives only if those points pass four tests: equal weight, no overlap, full coverage of the thesis, and a count small enough — two to four — for a human being to hold. Choosing a pattern comes next, and the choice is governed by the audience rather than the topic: chronological when the sequence is the argument, spatial when the structure is physical, causal when the link itself is disputed, problem-solution when the harm is already agreed upon, and Monroe’s motivated sequence when you need someone to actually do something. Structure that stays on the page does not exist, so transitions, signposts, internal previews, and internal summaries make it audible. The full-sentence outline is where the thinking gets tested, using coordination, subordination, the rule of division, and parallelism to expose weak logic while it is still cheap to fix. Then you compress that document into a keyword page with delivery cues, rehearse aloud from that exact page, and cut support rather than structure when the clock demands it.
Words worth knowing
| Term | What it means |
|---|---|
| Main point | One of the two to four top-level claims that together deliver the thesis; appears as a Roman numeral in the outline. |
| Organizational pattern | A reusable arrangement of main points (chronological, spatial, topical, causal, comparison, problem-solution) that carries its own built-in logic. |
| Chronological pattern | Points arranged by time or by the ordered steps of a process. |
| Spatial pattern | Points arranged by physical location, direction, or depth. |
| Topical pattern | Points arranged by the natural categories of the subject; the default when no other pattern fits. |
| Causal pattern | Points arranged as cause to effect, or effect back to cause; used when the link itself needs proving. |
| Problem-solution pattern | A structure that establishes a harm, often explains its cause, and then proposes a remedy. |
| Monroe’s motivated sequence | A five-step persuasive structure — attention, need, satisfaction, visualization, action — developed by Alan Monroe in the 1930s. |
| Connective | Any spoken device that makes structure audible: transitions, signposts, internal previews, internal summaries. |
| Transition | A bridge sentence that closes one point and opens the next by naming both. |
| Signpost | A brief verbal marker of position or emphasis, such as “second” or “the most important of these.” |
| Internal preview | A statement inside the body announcing the parts of the point you are about to develop. |
| Internal summary | A restatement, at the end of a point, of what that point established. |
| Full-sentence outline | The working or preparation outline in which every entry is a complete sentence, with citations and connectives written out. |
| Speaking outline | The keyword or delivery outline actually used while speaking; cues only, plus verbatim quotes and any statistic written with enough of its source to say it accurately. |
| Coordination | The principle that items at the same outline level carry roughly equal weight and scope. |
| Subordination | The principle that each indented item supports the item directly above it. |
| Rule of division | If an item is divided, it yields at least two parts: an A requires a B, a 1 requires a 2. |
| Parallelism | Sibling outline entries sharing grammatical form, each containing exactly one idea. |
| Chunking | Grouping small items into a few larger, labeled units so limited working memory can hold them. |