THE MACHINERY OF COMMUNICATION
A Complete Guide to Moving What You See Into Another Mind
The Unified Architecture of Transfer, Entry, Signal, Reading, Speaking, and Dialogue
What follows is not advice.
It is not a communication course. Not a leadership framework. Not a public speaking manual packaged in mechanism.
It is the complete architecture.
Six systems operating in parallel every time one person tries to make another person understand something. Transfer. Entry. Signal. Reading. Speaking. Dialogue. Each has its own machinery. Each runs its own computations. And all six must coordinate for communication to occur.
This document is the unified field. The practical integration of six mechanisms into a single operating system. With templates for how each mechanism deploys. With a phrasebook for common situations. With drills for building the coordination. And with the geometry that shows why these six and not others.
The mechanisms do not care about your personality. They do not care about your charisma. They care about structure. Get the structure right and the communication works. Get it wrong and no amount of eloquence saves it.
THE RECURRING PROBLEM
Two versions of the same moment.
A district manager needs to explain a new scheduling system to three area managers. The system changes how shifts are allocated. If the managers understand it, the transition takes a week. If they do not, it takes a month of confusion, resentment, and workarounds.
Version one. The manager opens a meeting. “So we are rolling out the new scheduling approach. Let me walk you through the details.” He shares his screen. Reads through the process document. Covers every edge case. Answers questions as they arise. The meeting runs forty-five minutes. Everyone says they understand. The next week, two of three managers implement it differently from each other and differently from what was described. The third calls to ask questions that were answered in the meeting.
Version two. The manager opens with: “Right now, you spend four hours a week fixing scheduling conflicts that should not exist. This changes that.” Pause. He sees the area managers lean forward. He continues with a single diagram showing the before and after. No edge cases. One model. He asks each manager to describe how they would handle next week’s schedule using the new system. One gets it right. Two get it partially right. He adjusts. Ten minutes later, all three can describe it. The meeting runs twenty minutes. The next week, all three implement it correctly.
Same content. Same manager. Same audience. Different architecture.
The difference is not talent. It is structure. Version two deployed all six mechanisms. Version one deployed none.
THE MECHANISM
Communication is not one skill. It is six systems coordinating.
THE SIX SYSTEMS OF COMMUNICATION
┌──────────────────────────────────────────────────┐
│ │
│ ENTRY ──→ SIGNAL ──→ TRANSFER │
│ │ │ │ │
│ │ │ │ │
│ ▼ ▼ ▼ │
│ SPEAKING READING DIALOGUE │
│ │
│ Top row: WHAT to communicate │
│ Bottom row: HOW to communicate │
│ │
│ Entry opens the gate. │
│ Signal compresses the payload. │
│ Transfer builds the model. │
│ Speaking encodes processing instructions. │
│ Reading closes the feedback loop. │
│ Dialogue enables joint construction. │
└──────────────────────────────────────────────────┘
Each system addresses a different failure mode:
-
Entry failure: The listener never engaged. The gate stayed shut. Content quality is irrelevant because the listener’s brain never allocated processing resources.
-
Signal failure: The listener engaged but was overwhelmed. Too much information. No hierarchy. The pattern was buried in noise. Working memory overflowed.
-
Transfer failure: The listener engaged with clean signal but built the wrong model. The anchors were misaligned. The abstraction level was wrong. The loading order skipped layers.
-
Speaking failure: The words were right but the delivery encoded conflicting information. The voice signaled uncertainty while the words claimed certainty. The pace was monotone. The processing instructions were absent.
-
Reading failure: The speaker never detected that the model was wrong. Open-loop communication. No verification. Confident misalignment.
-
Dialogue failure: Both parties exchanged information but nothing new emerged. Alternating monologues. No joint construction. Both left with what they came with.
Most people diagnose communication failure as “I was not clear enough” and respond by adding more words. The actual failure is almost always one of these six specific mechanisms. And adding words makes five of the six worse.
THE TEMPLATE
The geometry of communication is a loop with three phases.
THE COMMUNICATION LOOP
┌─────────────────────────────────────┐
│ │
│ PHASE 1: OPEN │
│ Entry + Stakes + Source Signal │
│ (first 7 seconds) │
│ │
│ Outcome: gate is open. │
│ Listener's brain has allocated │
│ processing resources. │
│ │
├─────────────────────────────────────┤
│ │
│ PHASE 2: BUILD │
│ Signal + Transfer + Speaking │
│ (the payload) │
│ │
│ Outcome: model is constructed. │
│ The pattern exists in the │
│ listener's architecture. │
│ │
├─────────────────────────────────────┤
│ │
│ PHASE 3: VERIFY │
│ Reading + Dialogue + Probe │
│ (the closed loop) │
│ │
│ Outcome: model is confirmed │
│ or corrected. The speaker knows │
│ what was built. │
│ │
└─────────────────────────────────────┘
Phase 3 feeds back into Phase 2.
The loop repeats until the model holds.
Phase 1 is the gate. Without it, Phase 2 is wasted. Phase 2 is the construction. Without it, Phase 3 has nothing to verify. Phase 3 is the closed loop. Without it, Phase 2 is unverified.
Most communication is Phase 2 only. No opening. No verification. Pure output into an unverified system. This is why most communication fails.
THE PHRASEBOOK
Opening (Entry Mechanism)
┌──────────────────────────────┐ ┌──────────────────────────────┐
│ CLOSED │ │ OPEN │
├──────────────────────────────┤ ├──────────────────────────────┤
│ │ │ │
│ "Let me walk you through │ │ "You are spending four │
│ the new process." │ │ hours a week on something │
│ │ │ that should take zero." │
│ "I want to talk about │ │ │
│ communication." │ │ "Every time you explain │
│ │ │ something and they nod but │
│ "Before I begin, let me │ │ do the wrong thing. There │
│ give some background." │ │ is a reason." │
│ │ │ │
│ "This is a complex topic │ │ "The best explainer you │
│ so bear with me." │ │ know does something you │
│ │ │ have never noticed." │
│ "Today's agenda covers │ │ │
│ three items." │ │ "What if clarity has │
│ │ │ nothing to do with how │
│ │ │ well you speak?" │
└──────────────────────────────┘ └──────────────────────────────┘
The closed column opens with the topic. The open column opens with the listener’s experience. The closed column gives context before the point. The open column gives the point before the context. The closed column signals “this will be long.” The open column signals “this matters to you right now.”
Building the Model (Transfer Mechanism)
┌──────────────────────────────┐ ┌──────────────────────────────┐
│ CLOSED │ │ OPEN │
├──────────────────────────────┤ ├──────────────────────────────┤
│ │ │ │
│ "Communication theory │ │ "Think of it like building │
│ dates back to Shannon's │ │ furniture in someone │
│ 1948 paper..." │ │ else's house. You have │
│ │ │ never been inside." │
│ "There are six key │ │ │
│ components to effective │ │ "This is not active │
│ communication..." │ │ listening. It is the │
│ │ │ opposite. Active │
│ "First, let me define │ │ constructing." │
│ what I mean by transfer, │ │ │
│ then signal, then..." │ │ "You already know how to │
│ │ │ read a room. This is the │
│ "To fully understand this │ │ mechanism underneath that │
│ you need to know about │ │ skill." │
│ working memory limits." │ │ │
│ │ │ "The simplest version: the │
│ │ │ listener builds, not │
│ │ │ receives." │
└──────────────────────────────┘ └──────────────────────────────┘
The closed column starts from the speaker’s architecture. The open column starts from the listener’s. The closed column provides background the listener did not ask for. The open column donates a model the listener can run. The closed column adds complexity. The open column compresses.
Compressing (Signal Mechanism)
┌──────────────────────────────┐ ┌──────────────────────────────┐
│ CLOSED │ │ OPEN │
├──────────────────────────────┤ ├──────────────────────────────┤
│ │ │ │
│ "There are several factors │ │ "Three things matter. │
│ to consider, including │ │ Everything else is noise." │
│ but not limited to..." │ │ │
│ │ │ "The constraint is working │
│ "While it is important to │ │ memory. Four items. That │
│ note that there are │ │ is the bottleneck." │
│ exceptions..." │ │ │
│ │ │ "Remove until the pattern │
│ "I should mention that │ │ is naked. Then stop." │
│ this does not apply in │ │ │
│ all cases..." │ │ "If you would not bet your │
│ │ │ salary on it mattering, │
│ "For context, the │ │ cut it." │
│ historical development │ │ │
│ of this approach..." │ │ │
└──────────────────────────────┘ └──────────────────────────────┘
The closed column hedges. The open column commits. The closed column adds qualifiers that protect the speaker. The open column removes everything that does not serve the listener. The closed column is precise. The open column is clear. These are different things.
Checking (Reading Mechanism)
┌──────────────────────────────┐ ┌──────────────────────────────┐
│ CLOSED │ │ OPEN │
├──────────────────────────────┤ ├──────────────────────────────┤
│ │ │ │
│ "Does that make sense?" │ │ "Explain it back to me as │
│ │ │ if I have never heard it." │
│ "Any questions?" │ │ │
│ │ │ "What would you do │
│ "Are we all on the │ │ differently on Monday │
│ same page?" │ │ based on this?" │
│ │ │ │
│ "Got it?" │ │ "Draw the process you just │
│ │ │ heard." │
│ "Okay?" │ │ │
│ │ │ "Where does this break?" │
│ │ │ │
│ │ │ "What would you tell your │
│ │ │ team this means for them?" │
└──────────────────────────────┘ └──────────────────────────────┘
The closed column asks for confirmation. The open column asks for construction. The closed column accepts “yes” as evidence. The open column requires output. The closed column verifies nothing. The open column reveals the model.
DOMAIN APPLICATIONS
Domain 1: Leading a Team
The manager explains a decision, a process, or a direction.
Entry: Start with what changes for them. Not the reasoning. Not the business context. What is different in their day starting Monday. “You will not fill out timesheets anymore. Here is what replaces it.”
Signal: One model. Not the five considerations that led to the decision. The one model that shows how their work changes. Everything else can be documented. The meeting is for the model, not the documentation.
Transfer: Anchor to their current process. “Right now you do X. The only thing changing is Y. Everything else stays.” Progressive loading: change one, check. Change two, check.
Speaking: Slow on the change itself. Fast on the context. Pause after the key shift. Let them process.
Reading: Watch for the furrowed brow (model conflict). Watch for the immediate nod (not processing). Ask one person to describe the new process. The gaps in their description are the gaps in your explanation.
Dialogue: “What am I not seeing about how this hits your team?” This is not a courtesy question. It is a construction act. They know things about implementation you do not. The meeting should produce a better model than the one you walked in with.
Domain 2: Teaching or Training
The instructor explains a concept, procedure, or framework.
Entry: Start with the problem the concept solves. Not the concept. The problem. “You have all been in a meeting where you explained something clearly and the other person did the opposite. This is why.”
Signal: Scaffold first. The simplest true version of the concept. Not the full version. The scaffold that everything else attaches to. “The listener does not receive understanding. They build it. From their own materials.”
Transfer: Progressive loading is everything in teaching. Layer one: the scaffold. Check. Layer two: the first mechanism. Check. Layer three: the constraint. Check. Never deliver layer three before layer two is verified.
Speaking: Variable pace. Slow on new concepts. Fast on familiar connections. Pause after each layer to allow integration. The pause is not silence. It is construction time.
Reading: Questions are your primary diagnostic. Clarifying questions mean the vocabulary is wrong. Structural questions mean the model is forming. Generative questions mean transfer is occurring. No questions at all means you need to probe.
Dialogue: “Based on what I just described, what would happen if…?” Run the model forward with the learners. Make them generate predictions from the architecture you built. If the predictions are wrong, you know exactly where the model diverged.
Domain 3: Persuading a Decision-Maker
The presenter needs a decision from someone with authority and limited time.
Entry: Stakes first. “This decision is costing us $X per week. I have a fix that takes one approval.” The decision-maker’s brain runs on cost-benefit computation. Feed it immediately.
Signal: Maximum compression. The decision-maker’s working memory is already loaded. You have capacity for one model, one recommendation, one number. Everything else is noise. “We switch from X to Y. Cost: Z. Timeline: W. Risk: one thing, here is the mitigation.”
Transfer: Model donation. Not a presentation. A thinking tool. “Think of it as moving from a toll road to a freeway. Same destination. No stops.” The decision-maker can now run the model on their own concerns without you narrating each one.
Speaking: Weight on the recommendation. Fast on the context. No hedging in the voice. The delivery must match the confidence of the recommendation.
Reading: Watch for the lean-back (evaluating, not rejecting). Watch for the specific question (engaged, testing your model). Watch for the redirect (“what about X?”= their actual concern, address it). Watch for the phone glance (gate closing, you lost them).
Dialogue: The decision-maker’s pushback is data. “What concerns you about this?” is not a polite question. It reveals the constraint you did not model. Integrate it. Adjust. The best pitches are not performances. They are dialogues that arrive at a shared model.
Domain 4: Difficult Conversations
One person needs to communicate something the other person does not want to hear.
Entry: Acknowledge the difficulty without preamble. “This is about your performance on the Henderson project. I want to be direct because you deserve that.” The gate opens on respect, not comfort.
Signal: One issue. Not three. Not “while we are at it.” One issue with one specific example. “The deliverable was due Thursday. It arrived Monday with errors that required two people to fix over the weekend.” No generalizations. No “you always.” One instance. Clear.
Transfer: Anchor to shared standards, not personal judgment. “The standard we agreed to is X. This fell short at Y.” The listener can evaluate the gap without evaluating the speaker’s fairness. The model is the standard, not the opinion.
Speaking: Even pace. Lower pitch. No emotional escalation in the voice even when describing emotional impact. The voice must signal “this is about the work, not about you as a person.” Weight on the standard. Lightness on the person.
Reading: The most critical reading moment in any communication. Watch for defensive posture (arms crossed, jaw set). This means the model is being received as an attack, not as a gap. Recalibrate. “I am not questioning your intent. I am showing you the gap between the target and where we landed.” Watch for the softening (they heard it). That is when the construction begins.
Dialogue: “What do you need from me to close this gap?” This converts the conversation from judgment to joint construction. The model shifts from “you failed” to “we have a gap. Let us build a bridge.”
Domain 5: Explaining to a Non-Expert
The expert explains a complex concept to someone outside the domain.
Entry: Start with what the concept does, not what it is. “This is the thing that decides whether your phone battery lasts eight hours or sixteen.” Not “lithium-ion energy density is determined by the cathode material composition.”
Signal: Ruthless subtraction. Everything the non-expert does not need to understand the operating principle is noise. They do not need the history. They do not need the edge cases. They do not need the qualifications. They need the model. One model. Clean.
Transfer: Model donation is the primary mechanism. “Think of the battery like a bucket of water with holes. The material determines the size of the holes. Smaller holes, slower drain, longer battery.” The non-expert now has a thinking tool. They can ask intelligent questions from it. “So what makes the holes bigger?”
Speaking: Slow. Simpler vocabulary delivered slowly is not condescending. It is respectful of the processing constraint. Pause after the model. Let it build. Do not rush to the next concept.
Reading: The non-expert’s face is your primary diagnostic. Glazed eyes = lost them, too abstract. Leaning forward = engaged, right level. Nodding too fast = performing understanding, not building it. The moment you see the glaze, stop. Back up. Find a better anchor.
Dialogue: “What does this remind you of in your own work?” The non-expert’s analogy reveals their anchor point. If the analogy is close, build from it. If it is misleading, gently redirect. “Close, but the difference is…” Their contribution tells you what scaffolding exists.
Domain 6: Writing (Asynchronous Communication)
The writer communicates without access to the reading loop.
Entry: The first sentence is the gate. Not the first paragraph. The first sentence. If it does not create a gap, declare stakes, or violate a prediction, the reader’s brain closes the tab.
Signal: Even more critical than in speech. Written words have no prosodic channel. No vocal weight. No pausing. The only hierarchy is structural: headings, paragraph position, sentence length. Short sentence after long paragraph = emphasis. Paragraph break = processing time.
Transfer: Every paragraph must earn its place. The reader’s working memory is the same bottleneck but without a speaker to pace the delivery. The reader controls the pace. If they feel overwhelmed, they stop. Scaffold first. One idea per paragraph. Progressive loading through document structure.
Speaking (adapted for text): Sentence rhythm replaces vocal rhythm. Short sentences slow the reader down (emphasis). Long sentences speed them up (context). Fragments stop them entirely (weight). The writer controls pace through syntax.
Reading (adapted for text): No real-time feedback loop. The writer must model the reader in advance. “Who is reading this? What do they already know? Where will they get confused?” Every editing pass is an attempt to close the loop prospectively. Cut everything the reader does not need to build the model.
Dialogue (adapted for text): Questions in the text serve as the writer’s side of the dialogue. “But what about X?” Anticipate the reader’s construction process. Answer the questions they would ask if they were in the room. The text becomes a simulated dialogue.
Domain 7: One-on-One Mentoring
The mentor shapes someone’s thinking over time.
Entry: Each session has its own gate. Do not assume the last session’s gate is still open. “What are you stuck on right now?” The mentee’s answer is the entry point. Not your agenda.
Signal: In mentoring, signal means asking more than telling. The mentor’s question compresses because it directs the mentee’s attention to the load-bearing element. “What is the one thing that, if you got it right, makes everything else easier?” One question. Maximum compression.
Transfer: Model donation is the mentor’s primary tool. Not giving answers. Giving thinking frameworks. “When I face this kind of decision, I ask two questions: What is the reversible version? And what information would change my mind?” The mentee now owns a tool they can use without the mentor.
Speaking: Match the mentee’s energy. If they are overwhelmed, slow and quiet. If they are energized, matched pace. The vocal channel in mentoring carries “I am here, I am with you, this is safe” more than it carries content.
Reading: Mentoring is the highest-bandwidth reading context. One-on-one. Extended time. Rich signal channel. Use it. Watch for the moment the mentee’s eyes change. That is the moment something connected. Build from there.
Dialogue: Mentoring is dialogue or it is lecture. The mentor who talks more than 40% of the time is not mentoring. They are presenting. “What do you think happens if you do that?” is worth more than five minutes of the mentor’s experience.
ERROR MODES
Error 1: The Knowledge Dump
Pattern: Speaker knows a lot about the topic. Pours it all out. Comprehensive. Thorough. Exhausting. The listener retains nothing.
Diagnosis: Signal failure. No hierarchy. No compression. The speaker confused being complete with being clear.
Fix: Before speaking, answer: “What are the three things that, if they understand nothing else, would still be useful?” Deliver those three. Stop. Check. If more is needed, deliver one more. Check again.
Error 2: The Echo Chamber
Pattern: Everyone nods. Everyone agrees. Everyone leaves. Nobody does anything different. The communication “went well.”
Diagnosis: Reading failure. The speaker trusted engagement signals instead of construction signals. No verification. No restatement. No behavioral check.
Fix: Replace “any questions?” with “tell me what you are going to do differently.” If the answer is vague, the model was not built. Go back to Phase 2.
Error 3: The Brilliant Monologue
Pattern: Speaker delivers a polished, articulate, impressive presentation. The audience is captivated. Q&A is light. A week later, nothing changed.
Diagnosis: Entry succeeded. Signal was decent. Transfer failed. The presentation was optimized for the speaker’s architecture, not the listener’s. Beautiful output into an incompatible system.
Fix: Start from the listener’s architecture. What do they already know? What is their current model? Build from there, not from your structure. The presentation that transfers is rarely the one that impresses.
THE DEEPER GEOMETRY
Why Six Systems, Not One
Communication is treated as a single skill because the output looks like one thing: a person talking. But the output is the surface. Beneath it, six independent systems are coordinating.
The reason they are independent is that they fail independently. You can have perfect signal and a closed gate (entry failure). You can have an open gate and overwhelming noise (signal failure). You can have clean signal that builds the wrong model (transfer failure). You can have the right model building with no verification (reading failure). You can have all four working but the voice undermining the words (speaking failure). You can have unidirectional excellence with zero emergence (dialogue failure).
Each failure requires its own diagnosis and its own repair. Telling someone to “communicate better” is like telling someone to “be healthier.” It is not wrong. It is not useful. The useful diagnosis names the failing system.
DIAGNOSTIC QUESTIONS
┌────────────────────────────────────────────────┐
│ Did they engage? → ENTRY │
│ Were they overwhelmed? → SIGNAL │
│ Did they build the wrong → TRANSFER │
│ model? │
│ Did the voice contradict → SPEAKING │
│ the words? │
│ Did I verify what they → READING │
│ built? │
│ Did we generate anything → DIALOGUE │
│ new together? │
└────────────────────────────────────────────────┘
The first "no" is the system to fix.
Fix that before touching anything else.
DRILLS
Drill 1: The Seven-Second Open
Practice: Before any explanation, write your opening sentence. Time constraint: it must be deliverable in seven seconds. It must create a gap, declare stakes, or violate a prediction. No preamble. No context. No “before we start.”
Test: Read the opening to someone unfamiliar with the topic. If they lean forward or ask a question, it works. If they wait politely, it does not.
Frequency: Before every meeting, presentation, or significant email for two weeks.
Drill 2: The Three-Element Compression
Practice: Take any explanation you need to give. Write it out fully. Then compress it to three elements. The three things the listener must understand. Everything else is cut. Not saved for later. Cut.
Test: Deliver only the three elements. If the listener can build a functional model from three elements alone, the compression was correct. If they cannot, one of the three was wrong.
Frequency: Daily for one month. Until compression becomes the default, not the exercise.
Drill 3: The Restatement Loop
Practice: After explaining anything to anyone, ask: “Tell me what you heard.” Not as a test. As a diagnostic. Listen to what they built. Identify the gaps. Identify the additions they made that you did not intend. Adjust.
Test: The gap between your model and their restatement shrinks over time. When the gap is consistently small, your transfer calibration is improving.
Frequency: Every explanation for one week. Then weekly maintenance.
Drill 4: The Silent Beat
Practice: In your next conversation, after every important point, stop. Count to three in your head. Do not fill the silence. Let it sit. Watch what happens to the other person’s face during the silence.
Test: You will see processing. The face changes. The eyes shift. They are building. If you had spoken, you would have interrupted the construction.
Frequency: Every conversation for one week. Until the pause becomes natural rather than forced.
Drill 5: The Dual-Channel Check
Practice: Record yourself explaining something. Play it back with your eyes closed. Listen only to the voice. Does the pace vary? Are there pauses? Does the pitch fall on conclusions? Does the voice carry weight on key points? Or is it monotone? Now read the transcript. Is the content clean? Compressed? Scaffolded?
Test: Both channels should carry information. If the voice is flat, the delivery layer is empty. If the content is noisy, the signal layer is overloaded. Fix whichever channel is weaker.
Frequency: Monthly. Record, review, adjust.
STACKING
Stack 1: Entry + Signal
Open with a compressed stakes declaration. Seven seconds. Three elements maximum. The entry IS the signal. No separation.
┌────────────────────────────────────────────┐
│ "Three things are broken in how we │
│ schedule. Each one costs four hours │
│ a week. Here is the fix." │
│ │
│ Entry: stakes declared. │
│ Signal: three things, specific cost. │
│ Gate opens. Noise is already cut. │
└────────────────────────────────────────────┘
Stack 2: Transfer + Speaking
Donate a model with vocal weight on the model itself. Fast on the context. Slow on the mechanism. Pause after the model lands.
┌────────────────────────────────────────────┐
│ (fast) "The old system makes you │
│ build the schedule from scratch." │
│ │
│ (slow, weighted) "The new system │
│ starts from last week's schedule │
│ and only changes what changed." │
│ │
│ (pause: 3 seconds) │
│ │
│ Transfer: anchor (old) + model (new) │
│ Speaking: pace + weight + pause │
└────────────────────────────────────────────┘
Stack 3: Reading + Dialogue
Verify what was built and use the verification to generate the next level of understanding.
┌────────────────────────────────────────────┐
│ "Walk me through how you would │
│ handle next week using this." │
│ │
│ Reading: their walkthrough reveals │
│ the model they built. │
│ │
│ "Good. Now what happens when you │
│ get a call-off on Tuesday?" │
│ │
│ Dialogue: generative question. │
│ They run the model on a new input. │
│ The answer reveals depth of transfer. │
└────────────────────────────────────────────┘
Stack 4: Signal + Reading (The Compression Check)
Deliver the compressed version. Then check what survived.
┌────────────────────────────────────────────┐
│ Deliver three elements. Nothing else. │
│ │
│ "If you had to tell someone who was │
│ not in this room, what would you say?" │
│ │
│ What they say back reveals what │
│ survived the compression. If element │
│ two dropped, the compression was wrong. │
│ Element two was not load-bearing. │
│ Replace it. │
└────────────────────────────────────────────┘
Stack 5: Entry + Transfer + Dialogue (The Full Loop)
Open the gate, build the model, generate together. The complete communication in its minimum viable form.
┌────────────────────────────────────────────┐
│ (Entry) "You are spending time on │
│ a problem that has a known solution." │
│ │
│ (Transfer) "Here is the model." │
│ [One diagram. One mechanism. One minute.] │
│ │
│ (Dialogue) "What am I missing about │
│ your specific situation?" │
│ │
│ The answer improves the model. │
│ The model improves the solution. │
│ Both leave with more than either │
│ brought. │
└────────────────────────────────────────────┘
WHAT IT IS NOT
Communication is not charisma. Charisma opens gates easily. But an open gate with no signal, no transfer, and no verification is an open gate to nothing.
Communication is not confidence. Confidence signals competence to the source evaluation circuit. But a trusted source delivering noise is still noise.
Communication is not vocabulary. Sophisticated language often increases the abstraction level beyond the listener’s architecture. The most effective communication uses the simplest words that carry the pattern.
Communication is not persuasion. Persuasion is one application of the six systems. The systems themselves are amoral. They build models. Whether the models are accurate or manipulative depends on the communicator’s intent, not on the machinery.
Communication is not a talent. It is six learnable systems. Each can be practiced independently. Each improves with repetition. The people who “communicate well” have internalized these systems through practice or accident. The systems do not require natural ability. They require structure.
THE SINGLE SENTENCE
┌────────────────────────────────────────────────────────┐
│ │
│ Open the gate. Compress the payload. │
│ Build from their materials. Verify what they built. │
│ Let the voice carry what the words cannot. │
│ Generate together what neither could alone. │
│ │
└────────────────────────────────────────────────────────┘
CITATIONS
Transfer and Reconstruction Bransford, J.D., Brown, A.L., & Cocking, R.R. (2000). How People Learn: Brain, Mind, Experience, and School. National Academies Press. Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences, 36(3), 181-204.
Entry and Attention Willis, J. & Todorov, A. (2006). First impressions: Making up your mind after a 100-ms exposure to a face. Psychological Science, 17(7), 592-598. Friston, K. (2010). The free-energy principle: A unified brain theory? Nature Reviews Neuroscience, 11(2), 127-138.
Signal and Working Memory Cowan, N. (2001). The magical number 4 in short-term memory. Behavioral and Brain Sciences, 24(1), 87-114. Sweller, J. (1988). Cognitive load during problem solving. Cognitive Science, 12(2), 257-285.
Reading and Feedback Ekman, P. (2003). Emotions Revealed. Times Books. Rozenblit, L. & Keil, F. (2002). The misunderstood limits of folk science: An illusion of explanatory depth. Cognitive Science, 26(5), 521-562.
Speaking and Prosody Zatorre, R.J. et al. (2002). Structure and function of auditory cortex: Music and speech. Trends in Cognitive Sciences, 6(1), 37-46. Schirmer, A. & Kotz, S.A. (2006). Beyond the right hemisphere: Brain mechanisms mediating vocal emotional processing. Trends in Cognitive Sciences, 10(1), 24-30.
Dialogue and Emergence Clark, H.H. (1996). Using Language. Cambridge University Press. Bohm, D. (1996). On Dialogue. Routledge. Sacks, H., Schegloff, E.A., & Jefferson, G. (1974). A simplest systematics for the organization of turn-taking for conversation. Language, 50(4), 696-735.
Mental Models Gentner, D. & Stevens, A.L. (1983). Mental Models. Lawrence Erlbaum Associates. Johnson-Laird, P.N. (1983). Mental Models. Harvard University Press.
Metacognition and Calibration Dunning, D. et al. (2003). Why people fail to recognize their own incompetence. Current Directions in Psychological Science, 12(3), 83-87. Koriat, A. (1997). Monitoring one’s own knowledge during study. Journal of Experimental Psychology: General, 126(4), 349-370.