Lesson 3, Note-Taking Systems

What to write, what to hold in your head, and the symbols that keep up with real-time speech.

First: when NOT to take notes

The most counter-intuitive rule in this course, straight from the source coaching:

Parts 1–3: put the pen down. Parts 4–6: you must write.

The early parts are short conversations, 1 to 1.5 minutes, two speakers, a contained situation. Notes there cost more than they earn: every second spent writing is a second of divided attention, and what these parts ask about most is tone and intent: which no pen can capture but full attention does.

The later parts are long, dense, single-play information streams, a news report, a three-person discussion, a three-minute viewpoint monologue. No memory holds all of that. Notes become mandatory.

One exception in the early parts: a rapid list - someone rattling off four features, three complaints, several prices, earns a quick grab of the pen even in Part 1. Lists are exactly what memory drops and paper keeps. Catch the list, put the pen back down.

The paper logistics (small thing, real marks)

You get one sheet of paper. When it’s full, you can ask for another, but staff take a minute or two to bring it, and the audio does not wait.

The rule: when your sheet is about three-quarters used, request the next one, during a gap, before you need it. Candidates have sat through the start of a Part 6 recording with a full sheet and no replacement. Entirely avoidable.

The column method

For any multi-speaker recording, structure beats speed. The system:

One column per speaker, positioned to match where they are.

In Part 5’s video, three people sit left, center, right on screen: your page gets a left, center, right column, one per person. In Part 6, each named speaker or organization gets a column. The spatial mapping isn’t cosmetic, when a question asks “what did the person in the middle say?”, your notes answer by location, instantly.

Three rules inside the columns:

1. Names go down first, always. Names are said fast, once, mid-sentence (“as Jonathan was saying-”), and the questions use them relentlessly: what did Clara believe? how did Ella react? A missed name orphans an entire column of notes. The moment you hear a name, it goes at the top of its column, spelled however sounds right.

2. Write top-to-bottom in conversation order, leaving gaps. When the left speaker talks, the left column fills; while she talks, the other columns stay blank at that height. This preserves sequence: and questions test sequence: who reacted to what? what did he say after she mentioned the cost? Your page becomes a timeline read down, a per-person summary read by column.

3. Arrows show who answered whom. A quick arrow from one column to another, he said this → she pushed back with that - captures the interaction structure that “who responded how” questions target.

The symbol toolkit

Real-time speech runs at ~150 words a minute; nobody writes that. Symbols close the gap:

Speed beats neatness, these notes live for ninety seconds. They exist to be glanced at during the questions, then abandoned. Ugly, fast, and structured wins over tidy, slow, and incomplete every time.

And the source’s blunt quantity rule: more notes generally wins. “If I write more than you, I win.” A candidate whose sheet is half-empty after Part 6 almost certainly under-captured. The symbol system exists precisely so more is physically possible.

Facts on paper, feelings in your head

The critical division of labor:

Write the facts. Hold the feelings.

You cannot write “she sounded hurt rather than angry, though she softened when he mentioned the refund” at speaking speed, and attempting it costs you the next two sentences. But feelings don’t need writing: tone, caught properly (Lesson 2), stays caught. Facts, numbers, names, sequences, lists, are what evaporate, and they’re cheap to write.

And whatever you choose to hold mentally, repeat it to yourself immediately: one silent echo (“hurt-not-angry, softened at refund”), or it’s gone by the next speaker turn.

The transcription trap

The failure mode of eager note-takers: writing without understanding.

The source’s example: a candidate hears a price, writes “75K,” and keeps transcribing. At question time, an option says “she was offended that he thought she wasn’t worth more than $75,000” - and the candidate, whose notes contain the number but not its meaning, matches the keyword and picks the trap. The actual situation, she was upset he’d spent that much without asking, was never in their notes because it was never in their processing.

Notes support understanding; they cannot replace it. If writing is crowding out comprehension, write less. A number you understand beats a page you don’t.

Worked example: what a good sheet looks like

A Part 5-style clip: three friends, Jonathan (center), Ella (left), Clara (right), discussing whether Jonathan should try high-intensity training.

   ELLA              JONATHAN            CLARA
   sarcastic ☺?      (center)            blunt, "reality check"
                     shy re: gym at 1st
   "you? HIIT? lol"  → defends: wants
                       to try smthg new
                     mentions Brian
                     James (athlete)
                                         ≠ warns: her cousin
                                         throws up after HIIT
                                         - only if > 1 hr !!
   ☺ warms up,
   offers to join
                     ☺ relieved

Notice: names first · columns match screen positions · gaps preserve who-spoke-when · the arrow marks a response · “only if > 1 hr” captures the qualifier - which is exactly what the trap option (“she said HIIT always makes people sick”) will test · feelings are one-word tags, not sentences.

Grade your own notes

After any Part 4–6 practice clip, audit the sheet:

• Every speaker’s name captured, at the top of a column
• Columns match speaker positions
• Readable as a timeline (down) and per person (across)
• At least five symbols/sketches doing the work of phrases
• Qualifiers caught, only if, unless, except, more than an hour
• Narrator’s words bracketed apart from speakers’
• Feelings held as one-word tags, not transcribed sentences
• I could still say what each person felt - proof writing didn’t eat listening

6+ ticks = the system is working. Below that, write less and structure more.

Your practice

Open one set on Celpora and put today’s lesson to work.

Practice Listening. Take notes in your own shorthand while the audio plays, then answer from your notes alone.