Outlining From Memory and Naming Your Knowledge Gaps
In the last lesson you dumped raw material onto the page without judging it. That dump is your raw ore. This lesson turns it into a skeleton you can trust, and, just as important, marks the exact places where your knowledge runs out, all before a search engine or an AI tool gets a turn. Knowing where your gaps are is what makes later AI help useful instead of invisible.
By the end of this lesson you can:
- Turn a freewrite into a memory-only outline with 3 to 5 main points.
- Sort every point into known, shaky, or unknown in under a minute.
- Rewrite a vague worry ("I don't know enough about this") into a named gap you can actually research.
- Keep a dated record of what came from your own head, which you will need at the capstone.
Build the skeleton before you look anything up
The rule is blunt: no browser, no notes, no AI, a timer set for eight minutes. The reason is not purity, it is information. The moment you open a source, your outline stops being a map of your mind and becomes a summary of someone else's. You lose the very thing this exercise produces.
Here is the sequence I ask writers to follow:
flowchart TD
A["📝<br/>Freewrite from 1.1"] --> B["🔁<br/>Circle repeated words"]
B --> C["🧱<br/>Skeleton: 3 to 5 points"]
C --> D["🚦<br/>Mark known / shaky / unknown"]
D --> E["❓<br/>Name the gaps"]
Four moves, in order:
- Read your freewrite once, slowly, pen in hand.
- Circle the words that keep coming back. Repeated words are your real interests, and they usually become your main points.
- Under each main point, list two to four sub-points from memory, in whatever broken phrasing arrives. Grammar is not the job yet.
- Leave a blank line wherever your mind goes quiet. A blank line is data, not failure.
Here is my memory-only outline on the topic "Is remote work good for new employees?" written in about eight minutes, markers included:
Sign in to view the full content of this lesson. Sign in
I. New hires learn a lot by accident (known)
- Side conversations and quick questions at a desk (known)
- The phrase "learning by osmosis" (shaky, I have heard it but cannot define it well)
II. Remote work removes most of those accidents (shaky)
- I believe this, but I cannot say how much or for whom (shaky)
- I have no number for how often remote new hires ask for help in week one (unknown)
III. What companies try instead (unknown)
- Buddy systems and scheduled check-ins: I have heard of both
- I could not describe either one accurately
Notice that the markers sit on sub-points as well as main points. Real outlines are mixed at every level.
Your turn (project thread). Take the freewrite you produced in lesson 1.1. Set an eight-minute timer and build the skeleton with no research. When the timer ends, count your blank lines and keep that number. It is your baseline for the module project.
Guidance if you got stuck
If your outline came out as one long list with no clusters, you probably circled too many words. Pick the three words that appeared most often and force everything else underneath them. If you circled none, read the freewrite aloud once; the words you repeat out loud are the ones to keep.
Marking each point: known, shaky, unknown
A marker is not a grade. It is a prediction about where you would get caught if a reader pushed back today. Use this test sheet:
| Marker | What it means | 20-second test |
|---|---|---|
| Known | I can explain it out loud without notes and give one concrete example | Say it to an empty chair |
| Shaky | I have the general idea but cannot defend a specific number, date, or causal claim | Ask "why exactly?" and listen for vagueness |
| Unknown | I recognize the term, but I could not say anything accurate about it | Try to define it in one sentence |
Two rules keep the system honest:
- When two markers both seem to fit, choose the weaker one. If you hesitate between known and shaky, it is shaky. Optimism here costs you later, when a claim collapses in revision.
- "Known" means confidence, not correctness. You will verify everything in lesson 3.1. For now you are mapping your own head, not certifying facts.
Your turn. Mark each line, then check yourself.
- "Remote work is common in software teams." ( )
- "Remote onboarding costs companies more money." ( )
- "Asynchronous communication means people reply on their own schedule." ( )
One defensible set of answers
Line 1 is known: you can explain it and give an example from a company you have heard of. Line 2 is shaky: it sounds plausible, but you cannot name a study, a cost figure, or a mechanism, so you cannot defend it under pressure. Line 3 is known only if you can restate it in your own words and give an example, such as a message thread answered across two time zones. If you were reciting the phrase without an example, mark it shaky. Your answers may differ from mine, and that is fine. What matters is that you can say why.
From vague worry to a named gap
Writers often write "I don't know enough about onboarding research" and then stall, because that sentence points nowhere. A named gap is specific enough to search for and specific enough to know when you are done.
The shape I use: I need to know ______ so I can ______.
- Vague: "I don't know enough about productivity research."
- Named: "I need to know of one study comparing how often new hires ask for help remotely versus in person, so I can support or drop my claim in section II."
Three gap types are worth tracking separately, because each needs a different kind of source:
- Fact gap: a number, date, name, or definition.
- Mechanism gap: how or why something works.
- Counterargument gap: the strongest objection someone would raise against you.
Write them into a gap log, not into the draft:
| Gap, written as a question | Type | Why I need it |
|---|---|---|
| How often do remote new hires ask for help in week one? | Fact | Decides whether claim II survives |
| What actually causes informal learning at a desk? | Mechanism | Lets me explain II instead of asserting it |
| Which studies argue remote onboarding works well? | Counterargument | I want the objection before I commit |
Your turn. Write three gap lines for your project, one of each type, using the "I need to know ___ so I can ___" shape. Keep them on a page you can hand in later. That page, plus your dated memory outline, is your evidence trail.
Rules of Thumb for an Honest Outline
- Timebox the memory pass at eight to ten minutes. Longer and you start inventing; shorter and you stop before the useful material surfaces.
- Do not fix anything while outlining. No polishing phrasing, no deleting the weird sub-point. The weird one is often where your real argument lives.
- Mark generously, at the sub-point level too. A main point can be known while every support under it is shaky. That is the most common and most useful thing an outline reveals.
- Write gaps as questions, never as topics. "Mentorship" is a topic. "What does mentorship look like when nobody shares a hallway?" is a gap.
- Keep your outline dated and unedited. When you use AI in Module 2, this file is the line between your thinking and the tool's contribution.
What to Carry Into Your Thesis
You now have three assets: a raw freewrite, a memory-based skeleton with markers, and a gap log with typed questions. Together they describe exactly what you can defend and exactly what you still owe your reader. That is the raw material a strong thesis is built from.
Next, in Thesis Development: From Topic to Defensible Claim, you will turn one of your main points into a single sentence you can argue, and use your gap log to test whether that sentence can survive the objections you have already predicted.
Front: What is a memory outline? Back: A rough outline built only from what you already hold in your head, before any research or AI help.
Front: How long should the memory-only outlining pass take? Back: About eight to ten minutes, timed, with no browsing.
Front: Which three markers does every outline point get? Back: Known, shaky, and unknown.
Front: What does a known marker mean? Back: You can explain the point out loud without notes and give one concrete example.
Front: What does a shaky marker mean? Back: You have the general idea but cannot defend a specific number, date, or causal claim.
Front: What does an unknown marker mean? Back: You recognize the term but could not say anything accurate about it.
Front: What is the quick test for a shaky point? Back: Ask why exactly and listen for vague phrasing.
Front: What do blank lines in a memory outline tell you? Back: Where your knowledge stops, which is exactly what you needed to find.
Front: When two markers both seem to fit, which do you pick? Back: The weaker one. Hesitation between known and shaky means shaky.
Front: What is the shape of a named knowledge gap? Back: I need to know ______ so I can ______.
Front: What are the three types of knowledge gap? Back: Fact gaps, mechanism gaps, and counterargument gaps.
Front: Why avoid research and AI during outlining? Back: The outline would become a summary of someone else's thinking, and your own map would be lost.
Front: Where do gap findings belong while you are still drafting? Back: In a separate gap log, not pasted into the draft yet.
Front: Does a known marker mean the point is accurate? Back: No. It records confidence only, and every point still needs verification later.
- A memory outline is built _____ any searching, browsing, or AI use, so the skeleton stays yours.
- A point marked _____ means you have the general idea but cannot defend a specific number, date, or causal claim.
- The three markers used on every outline point are _____, shaky, and _____.
- A named gap states what you need to know and _____.
- The three gap types are fact, mechanism, and _____.