How A-Heng Truly Learns Cognitive Psychology
A fictional learner's journey illustrates model testing, understanding, abstraction, transfer, and reliable performance.
Chinese original: 阿衡如何真正學懂認知心理學.
Editorial note: Full English translation of the user-supplied
knowledge.txt, added on September 28, 2026. The story is fictional, and its definitions, comparisons, and claims preserve the supplied author’s perspective. They have not been independently fact-checked and are not evidence of the learner’s mastery. The article title is supplied by the post metadata, and chapter headings are formatted for the site. The closing offer is retained as source text, not an instruction to create a training plan.
Related framing: Learning from Models to Performance.
The following story is fictional, but I will describe the internal cognitive processes of “someone with genuinely strong learning ability.” The point is not how many techniques he uses, but how, each time, he proposes models, exposes errors, revises models, abstracts structures, and transfers them to new problems.
Chapter 1: He Does Not Start Reading from Page One
At the start of his second year at university, A-Heng chooses a course in Cognitive Psychology.
The textbook is thick.
The first chapter introduces cognitive psychology. The later chapters cover perception, attention, memory, working memory, concepts, language, problem solving, judgment, and decision-making.
When his friend Xiao-Yu gets the book, the first thing he does is say:
“I’ll read 30 pages today.”
But A-Heng does not start reading.
He takes a blank sheet of paper and writes at the top:
“What exactly does cognitive psychology study?”
Before looking at the textbook at all, he tries to answer:
“Perhaps it studies how people receive information, process it, remember it, and use it to make judgments and take action.”
Then he asks:
“If I compressed the whole of cognitive psychology into a few important questions, what would they be?”
He provisionally writes:
How do people select information?
How does information enter memory?
How is memory retrieved?
Why do people make systematic errors?
How do people form concepts?
How do people solve unfamiliar problems?
Finally, he writes a line beside them:
These are only my current models. They may not be correct.
This action looks as though he has not “learned any content.”
But he has already started learning.
Because he is not preparing to move the textbook’s contents into his brain.
He is doing something more important first:
Building an initial model that can be revised.
This is the first difference between someone with strong learning ability and someone who simply works hard.
Xiao-Yu’s learning path is:
Textbook → memory.
A-Heng’s path is:
Prediction → learning → noticing discrepancies → model revision.
Without realizing it, he has already touched on “the essence of learning.”
Learning is not just adding information.
It is:
Continuously changing one’s internal model of the world in response to new experience.
Chapter 2: His First Real Understanding of “Knowledge”
In the first week, he learns about “working memory.”
The textbook explains that working memory has limited capacity, and that reasoning, reading, calculation, and other activities require people to temporarily maintain and manipulate information.
Xiao-Yu’s notes say:
“Working memory: limited capacity.”
“Attention: limited.”
“Long-term memory: very large capacity.”
None of this is wrong.
But A-Heng asks a different question:
“What exactly does limited capacity cause?”
He starts drawing relationships:
Lots of information
↓
Limited working memory
↓
Different pieces of information compete
↓
Selection is necessary
↓
Some information is retained
↓
Some information is ignored
He suddenly realizes:
“‘Working memory has limited capacity’ is not an isolated piece of knowledge.”
It can explain many phenomena.
Why can comprehension decline when you reply to messages while reading?
Why is it easy to make mistakes on a complicated mathematics problem when you must keep too many intermediate steps in mind at once?
Why do experts handle some problems more easily than novices?
Why can grouping information into larger meaningful units sometimes reduce the processing burden?
For the first time, a small model appears in his mind:
Limited processing resources + competing information → selection, forgetting, errors, and strategies.
At this point, he understands “the essence of knowledge.”
Knowledge is not:
“I remember that working memory has limited capacity.”
Knowledge begins to become:
I have built a structure that can explain different phenomena.
An isolated sentence is just a node.
Real knowledge consists of relationships formed between nodes.
His note-taking changes too.
He no longer mainly records:
“What A is.”
Instead, he starts recording:
Why does A occur?
What does A lead to?
What conditions does A depend on?
How are A and B related?
If A changes, what happens to B?
Information starts forming a network.
Knowledge starts forming a model.
Chapter 3: He First Discovers That “Following the Text” Is Not “Understanding”
A few weeks later, he studies long-term memory.
The textbook discusses encoding, storage, retrieval, retrieval practice, spaced learning, and related phenomena.
After reading a chapter, everything feels very smooth.
He understands almost every sentence.
So he closes the book.
He takes out a blank sheet of paper.
He writes:
“Without looking at the book, explain why testing can sometimes promote learning.”
He freezes.
Only scattered words come to mind:
“retrieval… memory… testing effect…”
But he cannot give a complete explanation.
For the first time, he clearly realizes:
Familiarity is not understanding.
While he was reading, the textbook was constantly supplying cues.
The sentences were right in front of him.
The concepts were right in front of him.
The author had already arranged the causal relationships.
So he developed an illusion:
“I know this.”
But when the external structure disappeared, he discovered that his own brain could not regenerate that structure.
That day, he sets himself a new “understanding test”:
To truly understand something, I should at least be able to regenerate it without the original text.
From then on, after each short section, he closes the material.
Instead of asking:
“Did I follow it?”
He asks:
“Can I reconstruct it myself?”
If he cannot, he does not call it “understanding.”
He only calls it:
“Something I have seen before.”
Chapter 4: Understanding Changes from “Answers” to “Models That Generate Answers”
Later, he studies the “spacing effect.”
If he were only memorizing, he could record:
“Distributed learning is usually more beneficial for long-term retention than massed learning.”
But A-Heng starts asking:
Why?
He reads different explanations.
Then he starts exploring counterfactuals.
What if two learning sessions had no interval at all?
What if the interval were so long that, by the second session, he had almost completely forgotten the material?
If an exam were only ten minutes away, would spaced learning necessarily have an advantage?
What if the goal were not long-term retention, but short-term familiarity with the material?
These questions matter.
Because he is moving from:
“Knowing a conclusion”
to:
“Knowing the conditions under which that conclusion holds.”
His understanding starts having four parts:
The phenomenon.
The mechanism.
The boundary conditions.
Predictions when the conditions change.
At this point, “understanding” truly appears for the first time.
Because his mind no longer contains only:
A causes B.
It begins to contain:
When conditions C and D are present, A affects B through mechanism M; if C changes, the result may change.
This kind of model can handle unfamiliar problems.
He gradually discovers:
The essence of understanding is not preserving answers someone else gives you, but building a model in your mind that can regenerate answers and predict changes.
Chapter 5: He Starts Truly Learning to “Abstract”
One day, A-Heng organizes the material from the previous weeks.
He notices several quite different phenomena.
Reading in a noisy café can be easily disrupted.
A demanding conversation while driving may increase cognitive load.
While doing complex mental arithmetic, remembering a string of digits becomes difficult.
Novices encountering complex concepts for the first time are more likely than experts to feel overloaded with information.
On the surface, these are completely different.
One is reading.
One is driving.
One is mathematics.
One is professional learning.
The old A-Heng would have recorded four phenomena.
Now he asks:
“If I remove all the surface elements, such as the café, the car, the numbers, and the textbook, what remains?”
He writes:
Limited cognitive resources facing multiple competing demands at the same time.
He pauses.
This is his first very clear act of abstraction.
He has not found a more impressive word.
He has simply removed irrelevant differences.
Reading, driving, mental arithmetic, and novice learning have all been stripped away.
What remains:
Limited capacity.
Competition.
Allocation.
Declining performance.
At this point, he understands:
Abstraction is not making language more mysterious, but removing what is unimportant and retaining relationships that still exist across different situations.
In other words:
Looking for invariants.
Chapter 6: But He Almost Falls into the Trap of “Over-Abstraction”
After getting the hang of abstraction, A-Heng becomes excited.
He finds that everything can be described as:
“Resources are limited.”
Later, he even thinks:
“Aren’t all problems in the world really about allocating resources?”
After hearing this, his teacher asks:
“If your model can explain everything, is there any phenomenon that could show it is wrong?”
A-Heng cannot answer.
The teacher says:
“Then perhaps you have not found a very deep model, but a very vague statement.”
For the first time, A-Heng realizes:
A higher level of abstraction does not mean greater explanatory power.
If a statement can explain any outcome, it may not really explain anything at all.
From then on, he adds a requirement for abstraction:
A good abstraction must not only encompass many phenomena, but also rule out certain possibilities.
In other words, it should generate predictions.
It should have boundaries.
There should be counterexamples.
His ability to abstract starts moving from:
Finding similarities
to:
Finding invariant structures with genuine explanatory power.
Chapter 7: The First Real Instance of “Transfer”
One evening, he is not studying psychology.
He is preparing an important classroom presentation.
Previously, he would have read his script ten times.
This time, he suddenly stops.
He realizes:
“Wait.”
He has studied retrieval.
Familiarity.
Contextual cues.
Working memory.
So he changes how he prepares.
He stops constantly looking at his script.
Instead, he closes the document, stands up, and speaks from memory.
When he gets stuck, he goes back to look.
Then he moves somewhere else and speaks again.
Finally, he asks a friend to pose random questions.
The next day, on stage, he forgets a phrase he had prepared.
Previously, this might have brought the whole passage to a halt.
But this time, he knows that what he really remembers is the whole argument’s structure, not a sequence of words.
So he quickly rephrases and continues.
At the end of the presentation, he suddenly understands:
This is transfer.
Cognitive psychology was originally knowledge in a textbook.
Now it has become:
A presentation strategy.
In two problems that look different on the surface, he sees the same structure:
When studying a textbook:
Retrieval → exposing gaps → revision.
When preparing a presentation:
Trying to explain aloud → exposing gaps → revision.
Therefore:
The essence of transfer is seeing an old, deeper structure beneath a new surface.
For the first time, he also understands:
Why the ability to abstract is a prerequisite for transfer.
If he had only remembered:
“Doing exam questions helps with exams.”
He might not have thought of it when preparing a presentation.
But if he abstracted it as:
Actively generating information from memory can expose one’s actual retrieval ability.
Then he might transfer it to:
Exams.
Presentations.
Interviews.
Teaching.
Writing.
Even speaking in meetings.
Chapter 8: Knowledge Starts Becoming “Ability”
After the midterm, the teacher stops setting purely conceptual questions.
Instead, the teacher gives a case:
A company wants to train employees to follow a new safety procedure.
The company has all employees watch eight hours of training videos in one day.
After training, everyone feels they have mastered the material.
But three weeks later, when actual problems arise, many employees cannot carry out the procedure correctly.
Analyze the possible causes and design a better training program.
Some students start trying to remember:
“Which chapter was this?”
A-Heng does not think of chapters first.
His mind starts working:
Eight hours of massed input.
A strong subjective sense of familiarity.
Very little active retrieval.
A long period without encountering the material again.
Differences between the actual situation and the learning situation.
What work requires is “retrieving the right procedure in response to the right cues,” not just “recognizing the video’s content.”
He starts building a model.
Then he proposes a plan:
Add distributed practice.
Add situational simulations.
Add active retrieval.
Add immediate feedback.
Test again after some time.
Finally, verify performance in the actual working situation.
What he is doing is no longer:
“Recalling cognitive psychology.”
It is:
Using cognitive psychology to solve a problem he has not encountered before.
This is when “ability” starts forming.
He finally understands:
Knowledge is a model.
Ability is whether you can reliably call on that model to produce results under real constraints.
Chapter 9: He Starts Learning to “Understand Regularities”
During a classroom discussion, a student says:
“I’ve noticed that students who study more than three hours a day usually get better grades. So studying three hours a day will definitely improve grades.”
Previously, A-Heng might have found this reasonable.
Now another set of questions automatically appears in his mind:
What exactly was observed?
Is this correlation or causation?
What other variables are involved?
Could people with stronger learning ability already be more willing to study for long periods?
Could motivation cause both “more study time” and “better grades”?
Could returns diminish beyond a certain amount of study time?
Is it the same across subjects?
Does “studying for three hours” involve the same content and methods?
A-Heng discovers:
Seeing two things change together and understanding the mechanism that generates that pattern are completely different things.
He gradually develops a regular habit.
When he sees a “regularity,” he neither immediately believes it nor immediately rejects it.
Instead, he looks for:
Variables.
Relationships.
Possible mechanisms.
Confounding factors.
Boundary conditions.
Evidence that can distinguish different explanations.
His ability to understand starts reaching another level:
Not memorizing regularities that other people summarize, but knowing how to judge what a regularity actually means.
Chapter 10: He Starts Learning to Understand “People”
Midway through the semester, they have a group assignment.
At the first meeting, a group member says:
“Everyone is fine with this plan, right?”
Nobody in the room objects.
Previously, A-Heng might have concluded:
“Everyone agrees.”
But now he does not.
Because he has learned to distinguish:
Observed facts
from:
His own interpretation of those facts.
What was actually observed was only:
One person proposed a plan.
The others did not publicly object.
What does “not objecting” mean?
They might really agree.
They might not care.
They might not have thought it through.
They might not want to be the first to object.
They might think objecting would be pointless.
They might fear slowing down the meeting.
A-Heng does not choose one explanation as “the truth.”
He keeps multiple hypotheses.
So he tries a different method.
He asks everyone to independently write down what they see as the plan’s biggest risk, without looking at anyone else’s answers.
As a result, six problems suddenly emerge in a plan to which “absolutely nobody had objected.”
That evening, he writes in his notes:
Understanding other people is not mind-reading.
It is building hypotheses from language, behavior, context, costs, incentives, and historical patterns, then letting new evidence update those hypotheses.
He adds:
Do not mistake the first plausible explanation for the only true motive.
This is where his ability to understand people starts maturing.
Chapter 11: He Then Discovers That Understanding “Groups” Is Harder
The group later encounters another problem.
Everyone says privately:
“We should start earlier.”
But whenever a deadline approaches, everyone still rushes through the work in the last two days.
An explanation based on individual personalities might be:
“Everyone procrastinates.”
But A-Heng starts asking:
“Why do different people start behaving similarly after joining this group?”
He finds:
Work is not clearly divided.
Nobody knows how far the others have progressed.
Finishing early has no obvious benefit.
Those who finish late only have to submit something at the end.
Everyone expects others might change their part, so they are reluctant to invest effort too early.
The real problem may not just be:
“All five people lack self-discipline.”
It may be:
The system is rewarding waiting.
A-Heng suggests:
Assign clear owners.
Make progress visible.
Split the final deadline into intermediate deliverables.
Expose dependencies earlier.
The group’s behavior really changes.
This time, he sees another, deeper layer:
Group outcomes sometimes cannot be inferred directly from each person’s individual intentions.
You must also study:
How information flows.
How power is distributed.
How responsibility is allocated.
What is rewarded.
What is punished.
What each person can see.
What each person expects others to do.
These structures can make people who are initially quite different behave similarly.
“Understanding people” and “understanding groups” become distinct in his mind.
Understanding an individual:
Person + situation.
Understanding a group:
People × people × rules × incentives × information × mutual expectations.
Chapter 12: He Starts Practicing “Understanding His Own Understanding”
As A-Heng continues studying, he encounters something strange.
On some questions, he is very confident.
But his answers are wrong.
On others, he is not particularly sure.
Yet he ends up being right.
So he starts doing something:
Before answering each question, he records:
How confident do I think I am?
For example:
90% certain.
70% certain.
50% certain.
After finishing, he checks the answers.
After a few weeks, he discovers an important problem.
His greatest danger is not:
“Not knowing.”
It is:
Not knowing that he does not know.
Some concepts are very familiar because he has seen them many times.
Familiarity brings high subjective confidence.
But when he is actually required to explain, derive, and apply them, the model is empty.
This is where metacognition starts forming.
Metacognition is not:
“I can think.”
It is:
Can I accurately estimate the quality of my own models?
From then on, his error notebook changes too.
Previously, it recorded:
“The correct answer is B.”
Now it records:
“Why did I choose A?”
Once, he writes:
My mistake was not just the answer.
I used an incorrect rule:
“More familiar = better learned.”
The correct model should distinguish “processing fluency” from “the ability to genuinely retrieve and apply something independently.”
At this moment, his errors start having value.
Because each error can expose a model.
Chapter 13: He Finally Understands What “Learning Ability” Is
Previously, A-Heng also believed:
People with strong learning ability are people who memorize quickly.
Now his understanding is completely different.
He sees his entire cycle:
Encounter something new.
Build an initial model first.
Generate predictions.
Learn.
Test himself.
Discover discrepancies.
Locate exactly which level has gone wrong.
Revise the model.
Test again.
Put the model into different situations.
Discover new boundaries.
Revise again.
So he writes a new definition of “learning ability”:
Learning ability is the efficiency with which someone uses limited experience to build models, discover errors in those models, revise them, and transfer them to new situations.
He finally understands:
A truly high-level learner is not someone who:
Makes fewer mistakes.
But someone for whom:
Each mistake leads to a higher-quality update.
Some students make ten mistakes and only remember ten correct answers.
Another student makes one mistake but discovers:
“So I have been using the wrong model X all along.”
The latter may eliminate dozens of future mistakes through one revision.
This is where differences in learning efficiency really open up.
Chapter 14: He Starts Practicing “Seeing Through to the Essence”
Before the final exam, A-Heng looks back over the whole cognitive psychology textbook.
He suddenly realizes:
If viewed only through the textbook’s chapters, it contains:
Attention.
Perception.
Memory.
Language.
Concepts.
Reasoning.
Decision-making.
Problem solving.
But viewed at another level of abstraction, many chapters revolve around recurring problems:
The amount of information the system can process is limited.
So selection is necessary.
Selection means some information is processed while other information is ignored.
The brain does not simply record the world passively; it interprets information under the influence of existing models, goals, and cues.
Memory is not simply a video recording, but involves encoding, retention, reconstruction, and retrieval.
Human judgment uses limited information and limited computational resources, so it uses shortcuts and, under some conditions, produces systematic biases.
Knowledge structures, in turn, affect new perception, attention, memory, and judgment.
Suddenly, a textbook of several hundred pages starts to “shrink.”
The content has not disappeared.
Instead, much of it can now be regenerated from a few structures.
For the first time, he truly experiences:
The deeper the understanding, the easier knowledge becomes to compress.
Because a beginner sees:
A hundred phenomena.
A more proficient learner starts seeing:
Ten regularities.
Someone who goes deeper sees:
A few structures that can generate those regularities.
This is “seeing through to the essence.”
Chapter 15: But What Counts as “Essence”?
A-Heng is not satisfied.
He asks:
“What exactly is the essence of cognitive psychology?”
At first, he writes:
Human cognition is information processing.
But he finds it too broad.
He writes again:
Human cognition is selecting, representing, retaining, retrieving, and using information under limited resources.
A little better than the previous sentence.
But he still knows:
This is only a useful generalization, not the universe’s uniquely correct ultimate statement.
He finally understands:
Essence is not a mysterious answer hidden beneath everything.
A more practical meaning is:
At the scale of the current problem, finding stable structures that can generate, constrain, and explain many surface phenomena.
Then:
What is the essence of essence?
A-Heng finally writes:
Finding invariants behind changes, and generative mechanisms behind phenomena.
In other words:
Invariance + generativity.
Chapter 16: The Last Question on the Final Exam
The teacher gives a final case that has never appeared in the textbook.
A company’s manager notices:
Employees score very highly every time they complete an online training quiz.
But months later, they still frequently make mistakes in actual work.
The manager thinks:
“These employees have a bad attitude. They did not pay attention during training at all.”
Analyze this.
A-Heng does not immediately accept the manager’s account.
Nor does he immediately reject it.
His whole cognitive system starts working.
Step one:
Distinguish phenomena from explanations.
The phenomena are:
High quiz scores.
Many errors in actual work.
“A bad attitude” is only one hypothesis.
Step two:
Look for key variables.
What is the quiz format?
How much time passes between training and work?
Does the quiz only involve recognition?
Does actual work require recognition or active retrieval?
Are the practice and work situations similar?
Is timely feedback available?
Step three:
Build multiple competing models.
It could be a motivation problem.
The training might produce only familiarity.
The quiz might be too easy.
The learning and application situations might be too different.
Time might lead to forgetting.
The workflow itself might induce errors.
Step four:
Look for evidence that can distinguish the models.
If it really is an attitude problem, would different training methods still make no difference?
If it is a retrieval problem, would adding uncued recall tests improve performance?
If it is a transfer problem, would adding realistic work simulations improve later performance?
If it is a system problem, would even people with high knowledge-test scores still make errors concentrated at certain points in the process?
Step five:
Redesign the system and predict the results.
He proposes:
Break up one-off training.
Add delayed testing.
Use active retrieval.
Add situational simulations.
Establish feedback in actual work.
Track where errors occur, not just quiz scores.
Then compare the actual effects of different plans.
The teacher did not ask:
“What is the definition of retrieval practice?”
But A-Heng called on almost everything he had learned that semester.
This is:
Knowledge → understanding → abstraction → transfer → ability.
And his entire analytical process is itself:
Thinking.
Chapter 17: Compressing A-Heng’s Growth over a Semester into One Sentence
At the start of the semester:
There are many blanks in his mind.
During the semester:
Those blanks are filled with large amounts of information.
Later:
Information starts forming relationships.
Relationships form models.
Common structures start appearing between different models.
Those common structures can transfer to new problems.
Repeated application develops stable ability.
Eventually, he no longer just:
Knows cognitive psychology.
He starts:
Thinking with cognitive psychology.
This is the change that really occurs in moving from “knowledge” to “ability.”
Chapter 18: Translating the Whole Story into a Learning Method You Can Use
Suppose tomorrow you begin studying a new field in depth.
Do not first pursue:
“How many pages will I read today?”
What you really need to build is this cycle:
| Stage | What A-Heng is doing | The question you should ask yourself | Ability being trained |
|---|---|---|---|
| Before exposure | Guessing first and building an initial model | What do I currently think is going on? | Thinking, metacognition |
| Input | Reading the actual content | What changed my original model? | Learning ability |
| Reconstruction | Closing the material and explaining again | How much can I still generate without cues? | Memory, understanding |
| Modeling | Drawing variables and relationships | What affects what? Why? | Ability to understand |
| Counterfactuals | Changing conditions | If X were absent, would the result still occur? | Causal understanding |
| Comparing multiple examples | Comparing different cases | What remains unchanged beneath different surfaces? | Ability to abstract |
| Counterexamples | Finding cases where the model fails | What looks similar but actually is not? | Awareness of boundaries |
| New situations | Solving unfamiliar problems | Which structure I have learned appears again? | Transfer ability |
| Real tasks | Actually completing a task under constraints | Can I reliably produce results? | Ability |
| Review | Analyzing why errors occurred | Was the answer wrong, or was my model wrong? | Learning ability |
| Calibration | Comparing confidence with results | Do I know what I do not know? | Metacognition |
| Further abstraction | Compressing many cases | What is the smallest set of principles that can generate the most phenomena? | Seeing through to the essence |
This table is what is really worth taking away from the whole story.
Chapter 19: How to Specifically Train the “Ability to Understand”
When learning a concept, do not stop at:
“What is its definition?”
For example, when learning “working memory.”
You should gradually push yourself to answer:
What are the key elements of working memory?
How are those elements related?
Why do capacity limits arise?
What phenomena can it explain?
What phenomena cannot be explained by it alone?
If its capacity suddenly became ten times larger, which behaviors might change?
How is it related to attention and long-term memory?
Can you use it to predict a situation you have never seen before?
When you can answer only the first question, what you mainly have is a definition.
As you gradually become able to answer the later questions, you start developing real understanding.
Chapter 20: How to Specifically Train the “Ability to Abstract”
Whenever you learn an important principle, deliberately find several cases that look very different on the surface.
For example, “limited resources.”
Do not only consider:
Students learning.
You can also compare:
Driving.
Meetings.
Multitasking at work.
Complex decision-making.
Then ask:
If you removed all the people, places, names, and technical terms, what would they still have in common?
Then ask the reverse:
Which differences cannot be removed?
Because the real difficulty of abstraction is not just knowing “what can be ignored.”
More importantly, it is:
Knowing what cannot be ignored.
Chapter 21: How to Specifically Train “Transfer Ability”
After learning a concept, deliberately stop doing only problems that look just like those in the textbook.
For example, after learning about retrieval practice.
Do not only ask:
“What does retrieval practice do?”
Instead, ask:
How can it be used in interviews?
In learning a foreign language?
In learning programming?
In presentations?
In employee training?
In doctors learning diagnosis?
If you find that you can use a principle only in its original problems, your knowledge is still tied to the surface situation.
Real transfer occurs when:
The shell changes, but you still recognize the skeleton.
Chapter 22: How to Improve the Ability to Understand People
The next time you see someone’s behavior, do not immediately write:
“They did it because of X.”
Change it to:
There are currently at least three possible explanations: X, Y, and Z.
Then ask:
What evidence supports X?
If Y were true, what else should I see?
What evidence could rule out Z?
Are their words consistent with their behavior?
What costs does their behavior require them to pay?
What constraints are they under?
What information do they have?
Most importantly:
Keep hypotheses open instead of rushing to judgment.
This will make your interpersonal understanding much more reliable than “intuitive mind-reading.”
Chapter 23: How to Improve the Ability to Understand Groups
When you see a group outcome, temporarily stop asking only:
“What personalities do these people have?”
Instead, ask:
If another set of people came in while the same institutional arrangements remained, would similar behavior occur again?
If the answer might be “yes,” then keep looking for:
Incentives.
Information flow.
Structures of authority and responsibility.
Rules.
Identity.
Mutual expectations.
Feedback loops.
This is moving from:
Personality-based explanation
to:
Structural explanation.
And this is an important step in seeing through complex social phenomena.
Chapter 24: How to Improve the Ability to “See Through to the Essence”
A-Heng eventually develops a thinking habit.
Whenever he encounters a complex problem, he digs downward in his mind:
Events
What happened?
↓
Patterns
What keeps recurring?
↓
Variables
Which factors are changing?
↓
Relationships
What changes together?
↓
Structure
How are these variables connected?
↓
Mechanism
Why does this structure produce this result?
↓
Boundaries
Under what conditions does it hold?
↓
Counterfactuals
If a factor were removed, would the result still occur?
↓
Predictions
If the model is true, what should happen next?
↓
Verification
Is reality actually like that?
↓
Revision
If not, where is my model wrong?
This whole path is the part of “seeing through to the essence” that can actually be trained.
Finally: Everything Is Really One Cycle
We can now compress the whole story to its core.
What A-Heng first sees is:
Large amounts of information.
He connects the information into:
Knowledge.
He discovers the causality and structure behind knowledge:
Understanding.
He extracts common invariant structures from different cases:
Abstraction.
He recognizes those structures again in unfamiliar problems:
Transfer.
He repeatedly uses them until he can reliably solve problems under real constraints:
Ability.
Throughout the process, he is constantly:
Comparing.
Decomposing.
Hypothesizing.
Reasoning.
Predicting.
Exploring counterfactuals.
Verifying.
Revising.
This is:
Thinking.
As he becomes increasingly skilled at completing the whole cycle:
Build a model → test the model → discover discrepancies → revise the model → abstract the model → transfer the model
His:
Learning ability
improves.
As he becomes increasingly skilled at building models that can explain, predict, and handle counterfactuals:
His ability to understand
improves.
As he becomes increasingly skilled at removing irrelevant differences from different phenomena and finding common relationships:
His ability to abstract
improves.
As he becomes increasingly skilled at looking beyond events and appearances to stable structures, generative mechanisms, and boundary conditions:
His ability to see through to the essence
improves.
So what you finally need to remember is not a dozen definitions.
It is this cognitive chain:
Phenomena
↓
Patterns
↓
Variables
↓
Relationships
↓
Structure
↓
Mechanism
↓
Abstraction
↓
Models
↓
Predictions
↓
Verification
↓
Revision
↓
Transfer
↓
Ability
If you keep learning along this chain over time, what changes is not just “how much you know.”
What really changes is:
How you form knowledge, understand the world, and solve unfamiliar problems.
And that is the deepest meaning of improving learning ability.
The most important point of this story is: Do not imitate A-Heng’s “study techniques”; imitate how he tests his models each time. Retrieval, counterexamples, analogies, varied cases, and review are effective because they all force your internal models to face the test of reality.
Going one level deeper, I could turn this story into a “practical training system”: using your actual self-study of cognitive psychology as the example, design exactly what to read, what to ask, how to reconstruct material with the book closed, how to train abstraction and transfer, and how to judge whether you have “memorized, understood, or genuinely developed ability” each day from Day 1 to Day 30.