From Experience to Understanding and Effective Action

Chinese source: 從經驗到理解與有效行動

Editorial provenance: Complete English translation of the user-supplied message.txt attachment dated 2026-09-29. Short section headings have been added for navigation; the Chinese counterpart preserves the attachment exactly after its source marker, with LF newline normalization only. This is translation, not fact-checking: the opaque ChatGPT citation markers are retained verbatim in their original order, but their missing bibliography remains unresolved and the cited claims have not been independently verified.

Related reading: Learning from Models to Performance · Knowledge, Reusable Structures, and Problem-Solving

A shared framework

The concepts you are asking about can be placed within a single framework: how people form knowledge and understanding from experience, how they judge whether they really understand, and how they turn that understanding into effective action.

Below, we will first clarify the concepts, then explain how to practise. The discussion of “essence” mainly takes a practical analytical perspective; philosophically, not all of these concepts have universally accepted final definitions.

What “essence” means

Clarifying “essence” first helps prevent the discussion that follows from becoming a game of words. In everyday talk, “looking at the essence” often mixes together three different questions:

What you really want to ask	What you are looking for	Example
What exactly is it?	The necessary properties that constitute it	A triangle is enclosed by three line segments; its colour and size can change
Why does it behave this way?	The mechanism that produces the phenomenon	If orders keep arriving faster than they can be completed, pending orders accumulate
What would be most useful to change now?	The current key constraint	Slow order preparation may be bottlenecked by the capacity of the only piece of equipment available

The answer to the third question is not necessarily the answer to the first. “What slowed down the work this time” is also not necessarily “the permanently unchanging essence of this kind of work.” Philosophical discussions of essence do indeed involve different understandings, such as “the properties that make a thing what it is” and “the properties it necessarily possesses.” :chatgpt-content-reference{index=”0”}

Thus, “What is the essence of essence?” can be understood as: on what grounds do we decide which properties and relationships are fundamental?

From the perspective of analytical method, the answer is: establish criteria for what to keep and what to leave out, distinguishing what cannot be omitted without preventing an accurate account of the object from details that can temporarily be ignored. This requires you first to make clear what you are explaining, at what level, and what evidence you have.

“Everything is about interests” and “Everything is energy” are brief, but they may not help you distinguish situations, predict outcomes, or choose actions. Making very broad generalizations and explaining things very precisely are two different things.

Distinguishing the concepts

The remaining concepts can initially be distinguished as follows. The “core” in this table is a generalization intended to help analysis, not a claim to have found the only philosophical answer.

Concept	What it is	A core you can grasp
Knowledge	A well-grounded grasp of facts, relationships, or ways of doing things	The connection between what you know and facts, evidence, and reliable methods
Learning	Relatively lasting changes in cognition, skills, or behavioural tendencies brought about by experience or practice	Experience changes how you subsequently judge and act
Understanding	A grasp of the relationships among a thing’s parts, why it is as it is, and when those relationships hold	Building a structure of relationships that supports explanation and inference
Abstraction	Retaining important features or relationships according to a purpose while temporarily ignoring other details	Leaving out details while preserving the structure you need
Transfer	Past learning influences new learning or tasks	The effect of previous experience in a new context; it may help, or it may interfere
Ability	The capacity to carry out a certain kind of task relatively reliably under certain conditions	Organizing and using knowledge, skills, strategies, and resources
Thinking	Forming and manipulating concepts, mental images, and relationships to judge, reason, imagine, and choose	Reorganizing information to explore relationships, possibilities, and consequences
Problem	A question awaiting an answer, or a gap between the current state and a goal that you do not yet know how to bridge	The relationships among unknowns, obstacles, goals, and constraints
Problem-solving	Clarifying a problem, choosing and carrying out methods, and checking whether an acceptable result has been obtained	Addressing gaps and uncertainty through reasoning and action

Two details need to be retained: learning can lead you to learn incorrect methods, so “learning has occurred” does not mean “correct knowledge has been acquired”; knowledge also cannot be judged solely by subjective certainty or getting an answer right by chance. Philosophically, even if a belief has reasons supporting it and happens to be true, that may still be insufficient for it to constitute knowledge. :chatgpt-content-reference{index=”1”}

One example connects them

Using the same example makes the relationships among these concepts clear.

Suppose a coffee shop receives ten orders per minute but can complete only eight, and this continues:

• You know how many orders arrive and how many are completed: you have acquired the relevant knowledge.
• You know why unfinished orders accumulate: you have developed understanding.
• You temporarily ignore the types of drinks and the shop’s decor, retaining only “inflow, processing, and backlog”: you have carried out abstraction.
• You use this structure to analyse software tasks or administrative approvals: you have carried out transfer.
• You can identify constraints, propose feasible improvements, and verify the results: you have demonstrated problem-solving ability.
• Through this process, you become able to analyse similar situations more accurately in the future: learning has occurred.

These processes can reinforce one another and do not necessarily occur in a fixed order.

Learning ability

Learning ability is the ability to acquire, retain, revise, and use new knowledge and skills. Its core is being able to adjust the learning process according to learning outcomes: knowing what you lack, choosing suitable methods, checking their effectiveness, and then making revisions.

You can begin improving it with the following five steps:

1. Write goals as behaviours you can demonstrate.
   “Understand a chapter” is difficult to verify; “explain the main concepts and complete a new problem without looking at notes” is clearer. When learning English, you can also write a goal as “actively use this phrase in three situations.”

2. Identify missing prerequisite knowledge.
   When you encounter material you cannot understand, circle the words, steps, or assumptions where you get stuck, then fill those gaps one by one. At the beginning, you can study a complete demonstration first, then gradually reduce the prompts.

3. Arrange retrieval without consulting materials.
   After closing the book, write down the key points, answer questions, or carry out the task, then check for errors. Research shows that, in particular scientific-text learning tasks, retrieval practice can improve performance on delayed tests, including questions requiring understanding and inference. :chatgpt-content-reference{index=”2”}

4. Spread practice across different days.
   You can initially try retrieving again the next day, a few days later, and a week later, then adjust according to what you forget. Spaced practice helps long-term retention, but no single schedule suits all material and goals. :chatgpt-content-reference{index=”3”}

5. Choose the next step based on errors.
   If you cannot understand, you need to fill in concepts or change examples; if you understand but cannot recall, you need retrieval practice; if you can explain but cannot do it, you need hands-on practice; if you can solve familiar problems but not new ones, you need to compare different contexts.

Here is a very useful shift: rewrite “My learning ability is poor” as “Am I currently stuck at understanding, memory, retrieval, execution, or transfer?” The latter question points directly toward practice.

Understanding ability

Understanding ability is the ability to build, check, and revise models of relationships among things. Here, a “model” is your simplified explanation of “which factors influence one another, and how.”

Its core is enabling you to infer from what is known something you have not been directly told.

Whenever you learn a concept, you can check it with six questions:

• Meaning: Can you explain clearly in your own words what it refers to?
• Relationships: Which known concepts does it connect with? Where do they differ?
• Reasons: Why does this outcome occur? What steps take place in between?
• Examples: Can you come up with an example and a counterexample yourself?
• Changes: If one condition changes, how might the result change?
• Boundaries: Under what circumstances does it not hold, or need modification?

For example, if you say “Adding staff can shorten waiting times in a coffee shop,” you should then ask: if the real limit on output is the only coffee machine, what effect would adding staff have?

Handling this kind of change in conditions demonstrates understanding better than simply repeating “Adding staff improves efficiency.” Verbal explanation is only one way to check; calculations, diagrams, hands-on execution, and solving new problems can also test understanding.

Understanding regularities

The ability to understand regularities is the ability to identify stable, conditional relationships from individual events and check whether they can be applied to new situations. It requires grasping both “what repeats” and “when there are exceptions.”

You can organize a vague statement into this form:

Under condition C, A increases or decreases B through mechanism M; if condition D arises, the relationship may change.

For example, “There are lots of people, so service is poor” is too vague. A more testable statement is:

When orders continually exceed processing capacity, and there is no cancellation or diversion, unfinished orders accumulate and waiting times may increase.

There are three things to do when practising:

1. Look for cases where the relationship holds and where it does not. Do not collect only stories that support your existing view.
2. Compare other explanations. If A and B occur together, A might affect B, or both might be affected by C.
3. Make predictions before looking at new data. Write down “If my understanding is correct, what should I see next?” to avoid always being able to invent an explanation after the fact.

Some regularities are strict relationships; others are only probabilistic trends. Understanding regularities also includes knowing which kind you are using.

Understanding people and groups

Understanding other people and groups means forming and revising judgments about their meanings, goals, and constraints based on words, behaviour, and context. Words and behaviour can be observed; inner thoughts usually can only be inferred indirectly.

When analysing an individual, you can follow four steps in sequence:

1. Record what was actually said and done.
   “He did not answer this question” is an observable description; “He looks down on me” is an interpretation.

2. Add the context.
   What information did the other person have? What pressures were they facing at the time? How much time and what resources did they have? What did they value? What emotions, habits, and responsibilities were involved?

3. Retain multiple reasonable hypotheses.
   “Decide for yourself” could express trust, busyness, delegation of authority, or dissatisfaction. You need to judge it in its surrounding context.

4. Check through conversation.
   You can ask “What matters most to you right now?” or “I understand you to mean …, is that right?” A set of studies found that merely imagining the other person’s perspective did not consistently improve judgment accuracy; experiments within that set that obtained information directly through conversation did improve accuracy. :chatgpt-content-reference{index=”4”}

When analysing a group, you also need to include decision rules and differences among members:

• Who can decide? Who can veto?
• Who benefits? Who bears the costs?
• What information does each member have?
• How are rewards, evaluations, and responsibilities allocated?
• Does the public position represent every member’s views?

For example, suppose an organization verbally encourages innovation, but the responsibility for failure falls primarily on the person making the proposal. Members may then tend to be conservative. This is an institutional explanation worth checking, but it still requires evidence for confirmation.

A collective outcome may emerge from the interaction of different positions, without anyone fully intending that outcome. Thus, understanding a group requires looking at individual motives, modes of interaction, and institutions together.

Abstraction and transfer

Abstraction ability is the ability to identify and represent important features and relationships according to a purpose. Its core is knowing “which details must be retained in order to answer this question.”

You can practise in four steps:

1. Determine the purpose first. Do you want to classify, explain causes, predict outcomes, or design a method?
2. Compare concrete cases. Find situations that differ on the surface but have similar relationships.
3. Replace names with roles and relationships. For example, replace “barista” with “processing resource.”
4. Return to the concrete cases to check. See whether the abstraction has left out important differences that would change the conclusion.

For example:

Setting	Units that need processing	Processing resources	Unfinished work
Coffee shop	Orders	Staff and equipment	Waiting orders
Software system	Tasks	Processing resources	Task queue
Administrative approval	Applications	Approval staff	Pending applications

What you extract from these is the relationship among “work inflow, processing capacity, and work backlog.” “They are all systems” makes a very broad classification; identifying specific relationships further is more helpful for analysis.

Comparing cases to identify a shared structure also has research support: in research on learning negotiation, directly comparing cases can promote the extraction of principles and their subsequent use. This supports including “side-by-side comparison” in practice, but does not justify assuming that any knowledge will automatically transfer across domains. :chatgpt-content-reference{index=”5”}

Transfer ability, in turn, requires checking how far this abstract structure actually applies in a new context. When practising, besides finding similarities, write down at least one important difference.

For example, customers may abandon a queue; whether software tasks will be cancelled or time out depends on the particular design. This difference may change the outcome of the backlog. Similarity provides a starting point for analysis, but verification in the new context is still needed.

Problem-solving ability

Problem-solving ability is the ability to identify the gap that genuinely needs addressing, choose methods, organize resources, and adjust according to the results. Its core is connecting “problem representation, reasoning, action, and feedback.”

You can use the following process:

1. Define success. What outcome would count as an improvement? Who judges it? What constraints are there?
2. Distinguish what is known, what is unknown, and what is assumed. Label guesses so that they are not treated as facts.
3. Break down the obstacles. Is information, skill, or resources missing? Has collaboration failed, or do the goals conflict with one another?
4. Propose several candidate explanations or methods. Compare costs, effectiveness, feasibility, and trade-offs.
5. Choose a small action that can provide information. Prioritize trials that distinguish among hypotheses and are easy to adjust.
6. Check the results and revise. Record which judgments held, which need changing, and the conditions under which the method applies.

For example, “I can’t learn English” can be rewritten as:

I can recognize some phrases, but in conversations without prompts I often cannot recall them.

This lets you distinguish further: is the meaning unclear, is retrieval slow, are connections to contexts missing, or is more actual conversation practice needed? Each obstacle calls for a different kind of practice.

When improving problem-solving ability, besides practising the completion of tasks, you can also reconstruct your judgment process afterwards: what information was available at the time? Why did you choose this method? Which outcomes were unexpected? What could you notice earlier next time? At the same time, you need to accumulate knowledge in the relevant domain; general methods require concrete content to be effective.

Seeing the essence more clearly

To improve your ability to “see through to the essence,” you can set your goal as finding an explanation that is supported by evidence, explains key differences, and has known boundaries of applicability.

Whenever you analyse something, write down these six items:

What to write down	What it checks
The specific phenomenon I am explaining	Whether the problem is clear
Known facts and their sources	Whether impressions are being treated as facts
Factors and relationships that may be at work	Whether the account stops at description
At least one competing explanation	Whether an answer has been settled on too early
Observations or actions that distinguish these explanations	Whether the account can be tested
Conditions of applicability and possible counterexamples	Whether the account overgeneralizes

For example, “He procrastinates because he is lazy” provides little information. You can check further: is the task vague? Does he lack the skills? Is he waiting for information from someone else? Is he afraid of being evaluated after delivery? These explanations predict different behaviours and lead to different methods of improvement.

Analysis does not require asking “why” endlessly. When you already have enough evidence to make the current decision, and further questioning is unlikely to change your choice, you can act first, then use the results to update your understanding. Different tasks require different depths of explanation.

A daily practice loop

To begin practising, you can start by spending thirty minutes each day completing the following process around one small concept or a real problem. The times are merely a convenient starting point for putting it into practice and can be adjusted according to difficulty.

Time	What to do
5 minutes	Without consulting materials, recall yesterday’s learning or write down the known conditions of today’s problem
8 minutes	Learn a small concept, study a demonstration, or fill in one key piece of information
5 minutes	Explain the relationships in your own words and propose an example and a counterexample
8 minutes	Switch contexts or change one condition, then try to make a prediction or solve a problem
4 minutes	Check the results and record the reasons for errors and the question to verify next time

After two weeks, use problems or situations you have not practised to check: can you still retrieve what you learned after a few days? Can you reason when conditions change? Can you identify the parts you are uncertain about? After discovering an error, can you choose a more suitable next step?

You can start today with “understanding,” which you raised: without looking at the material above, use an example of your own to explain the difference between understanding and remembering, then change one condition in the example and see whether you can infer the outcome.