Skip to content
C.W.K.
Stream
Lesson 02 of 05 · published

Four Things Speech Structurally Cannot Do

~13 min · visual-design, information, argument, primary-sources

Level 0Cold Workshop
0 XP0/35 lessons0/12 achievements
0/100 XP to next level100 XP to go0% complete

The Positive Half of the Rule

"Never repeat the narration" tells you what not to put on screen. It does not tell you what to put there. The positive half is a list of four capabilities, chosen because each one is something speech cannot do — not something it does badly, something it structurally cannot do at all.

1. Simultaneity

Speech is linear. It delivers one clause at a time and the listener has to hold the earlier ones in memory, which they will not reliably do. A picture can hold the entire argument at once. So the screen accumulates structure across the piece — each beat adds a node, nothing is erased — and by the end the whole shape is present simultaneously. That final frame is also the natural source for a thumbnail, because it is the one image that contains the argument.

2. Visible Revision

Narration can say "I was wrong about that." Only the screen can show the correction happen — a node that was there gets struck, an annotation gets erased, a claim re-reads differently once something beside it changes. This is the signature beat of the format, and it is worth understanding why it is so much stronger visually. Spoken, a correction is one more sentence in a stream of sentences. Shown, it is a change to something the viewer has been looking at, which means they witness it rather than being told about it.

3. Quantity as Shape

A sequence of numbers read aloud is a list, and a list of four numbers is already more than a listener will hold. Plotted on an axis, the same four numbers are a shape — and the shape is usually the actual point. An oscillation, a plateau, a sudden jump: none of those survive being spoken, and all of them are instant when drawn.

4. Primary Sources

The last one is not about perception, it is about possession. When the source material carries real artifacts — an attachment, a screenshot, a document someone actually sent — those are things nobody else has. Showing the real thing is categorically different from describing it, and it is the one category of screen content that cannot be reproduced by a competitor with a better narrator.

Pick channels by what each one uniquely enables. The four capabilities above are not a style guide; they are the answer to "what is a picture for, given that a voice is already running?" Ask the same question of any second channel you have — a log beside a metric, a diagram beside prose — and the useful content is whatever the primary channel structurally cannot express.

Code

The accumulating argument, three beats in·text
beat 3                      beat 6                      beat 9

  [claim A]                   [claim A]                   [claim A]
                                  │                           │
                              [claim B]                   [claim B]
                                                              │
                                                          [claim C]

# nothing is erased between beats - the viewer who looks up
# at beat 9 can still see what was established at beat 3.
# speech cannot do this: by beat 9 the words of beat 3 are gone.

# the revision beat, which is the one speech cannot show at all:

  [claim A]                   [claim A]
      │           ───▶            │
  [annotation B]              [~~annotation B~~]   <- struck, ON SCREEN
                                  │
                              [claim B']          <- the corrected reading

External links

Exercise

Take an explanation you have given recently — a document, a talk, a long message — and identify one moment where you corrected an earlier claim. Design how that correction would look if it were shown rather than said: what was on screen before, what changes, and what stays. Then ask whether the visual version would have landed harder, and why.
Hint
The strength comes from the viewer having already accepted the thing being struck. A correction shown at the moment of correction works; the same information presented as "here is the right answer, and by the way people used to think otherwise" is a fact rather than an event, and it does not land the same way.

Progress

Progress is local-only — sign in to sync across devices.
Spotted a bug or have feedback on this page?Report an Issue

Comments 0

🔔 Reply notifications (sign in)
Sign inPlease sign in to comment.

No comments yet — be the first.