How Shushu Works: Professional Calculation, Interpretation, and System Design
Shushu models can only represent selected variables and relationships within a complex reality. The results of professional calculation therefore need to be considered in relation to the particular people and circumstances involved, and refined through observation, comparison, and feedback. In Shushu, the diviner is the reasoning subject. Professional calculation, interpretation, source retrieval, and feedback from real circumstances all contribute to an ongoing process of reasoning, within which the diviner forms judgments and chooses how to act within the system's boundaries.
Wang Chong’s Lunheng, in the chapter “Shizhi” (“Actual Knowledge”), tells of a disciple of Zhan He who heard an ox lowing outside the gate and inferred that it was black, with white hooves. Zhan He agreed and sent someone to check. The ox was indeed black, its hooves wrapped in white cloth.
If one can infer such things through shushu—reading signs in time and reasoning from hexagrams—why should a sage not be able to know a thousand years ahead and ten thousand ages into the future?
If you can tell that an ox is black and its hooves are white, do you also know who owns it—or why its hooves were whitened?
Shushu models can reveal only one aspect of a matter; they cannot capture its full reality. Even when we examine a matter closely and question it from every angle, much may still elude us. Why? Because without seeing it with our own eyes and asking about it directly, we cannot know it fully.[1]
Shushu has limits because its calculations select certain variables from a complex reality and show how they appear under particular rules. They cannot account for every cause behind what we observe. Does that make shushu false or ineffective? No. Shushu can produce valid judgments within a defined scope, but their validity remains bounded by the inputs, the rules of calculation, and the range to which those rules apply.
In Shushu, the diviner is the one who reasons and performs the calculations. Supported by underlying large language models and algorithmic engines, diviners have clear advantages in both knowledge and computational capacity. Yet a particular person or question always belongs to a reality full of uncertainty. The more important task is to keep comparing and calibrating the results of shushu calculations against the complexity of real life, and to decide from what perspective the available information should be understood.
1. Reasoning from the Unknown
A person may enter a conversation with incomplete birth information:
I know my date of birth and birthplace, but I’m not sure of my exact birth time. What should I do?1.1 Establishing the Prerequisites
The diviner begins by checking the available birth details and calibrating the time information: the date and place of birth, sex, and whether the person can recall an approximate time range. This is enough to form an initial assessment and decide what to ask next. For example:
When something happens, are you more inclined to observe first and decide afterward, or do you usually act first and adjust as you go?Or:
Have you ever experienced a major, long-term move, such as relocating to another city or country? If so, roughly which year was it?Questions like these help narrow the possibilities. The diviner compares the candidate charts against the answers, gradually reducing the range of possible birth times. After several rounds of exchange, only a small number of better-supported candidates remain for more complete professional calculations.
Birth-time calibration draws on two main kinds of evidence. One is long-term state evidence: personal characteristics, behavioral patterns, or relationship conditions that persist over time. The other is event-time evidence: changes in real life that occur in a particular year, month, or on a specific date. The diviner checks both kinds against the differences shown by the candidate charts.
Time calibration → initial check against long-term patterns → candidate birth times → candidate chart calculations → questions focused on points of difference → comparison with the timing of life events → narrowing the range
The remaining candidates are then processed separately through the full professional calculation pipeline, producing more complete charts and time-calibration results. These are further compared across BaZi, Zi Wei Dou Shu, Qizheng Siyu, and other professional materials relevant to the case. As the evidence converges, the diviner proposes the working birth time that currently has the strongest support. Once the user confirms it, the confirmed birth time becomes the input for subsequent formal chart calculations.
1.2 Developing Interpretive Material
A chart produced through more comprehensive professional calculations provides richer material for analysis and makes it possible to ask more specific questions. The diviner can now use different shushu tools to examine finer details. For example:
I feel lost about the future. I don’t like my current job, but I don’t know what kind of work would suit me either. What makes it even harder is that I can’t tell which of the two possible birth times describes me more accurately.The diviner first runs the relevant professional calculations and compares the two candidate birth times. Looking at career, long-term personal tendencies, and changes in the current stage of life, the diviner identifies structural differences that are more useful for distinguishing between the candidates. These professional results are then reorganized into differences the user can recognize in their own circumstances.
The diviner might explain that one candidate is more consistent with someone who has long felt disconnected from their current career direction, with changes pointing toward a new career path. The other may be more consistent with someone who still retains the abilities and experience built up in the existing field, but whose current way of working and environment have become difficult to sustain. The user does not need to understand the chart itself; they only need to judge which description comes closer to their actual circumstances.
1.3 Feedback Refines Judgment
It’s hard for me to resign or switch industries. I feel trapped. But I actually don’t want to stay in this industry. I want to become self-employed, but my real circumstances don’t allow it.This new feedback enters the next round of information screening and judgment, giving the question “What kind of work suits me?” a deeper basis for analysis.
Professional calculation → comparison of candidates → new observations from multiple shushu models → recognizable differences in real-life circumstances → user feedback → renewed screening and judgment → further action guided by the new information gap
The user’s emotions, work situation, and practical constraints all provide factual information against which the diviner can calibrate. New facts may strengthen one candidate or expose weaknesses in an earlier judgment. The diviner must reorganize the professional charts already obtained, consult relevant classical texts, and incorporate the user’s feedback before deciding what to do next: continue comparing candidates, bring multiple shushu models into a combined analysis, or clarify a key condition in the user’s actual circumstances.
A single inquiry can therefore develop into an ongoing cycle of feedback:
Current understanding → action → new material or results → update understanding when the basis for judgment has materially changed → revise the provisional judgment → decide what to do next
As the user provides new information about real life, the diviner continues to call on professional calculations, compare results, and consult relevant sources. The process moves from having no idea of the birth time, to identifying a small number of candidates, to calibrating them step by step and settling on the one that is currently “most likely.” The judgment ultimately returns to the user’s actual circumstances.
This progression from the unknown toward a more specific judgment relies on the professional observations and calculations offered by shushu models, while remaining subject to calibration against real-life information and subsequent feedback. Shushu models organize observation through time, space, position, relations, and change, then place particular people and events within those structures. A person’s circumstances must therefore first be situated within clear relations of time and space. Only then do calculation, comparison, and calibration have a common frame of reference.
The first necessary condition for Shushu’s professional reasoning, then, is to establish time, space, and the subject being examined.
2. Anchoring Time and the Subject
Shushu analysis begins with a more basic condition: people and events always occur at particular times and places. When a shushu system uses these conditions as inputs, its implementation must place them consistently within the relevant time-and-space structure.
2.1 How Time, Space, and the Subject Become Inputs
A shushu model that relies on the time and place of birth or inquiry as key inputs can be understood as an algorithmic structure built on astronomy, calendrical systems, and spatiotemporal relations. It does not reproduce reality in its entirety. Instead, it selects a limited set of variables and organizes celestial motion, cycles, positions, and the relations among them into parameters that can be calculated. A particular person or event is then situated within this structure for analysis.
For these forms of shushu, precise time is one of the most basic inputs to professional calculation. Celestial bodies are in constant motion, and Earth rotates on its axis and revolves around the Sun. A specific moment corresponds to specific positions within these cycles, while the observer’s location on Earth determines the relationship between celestial bodies and local space. Once time and location have been established, a concrete moment in reality can be converted accurately, according to the rules of the relevant shushu model, into the time, position, and other parameters required for calculation. The subject can then be located within the corresponding spatiotemporal structure.
What these models analyze is the situation of a particular subject within this set of astronomical, calendrical, and spatiotemporal relations. If the time is inaccurate, the subject is placed within a different and incorrect set of relations, leaving the resulting calculation without a sound basis. Accurate spatiotemporal inputs are therefore a prerequisite for this kind of professional calculation.
Astronomical and calendrical structures → precise spatiotemporal inputs → locating the subject → establishing calculable circumstances
2.2 Different Coordinates of Time
Once time enters professional calculation, several distinct concepts come into play. They do not all describe the same thing: some refer to the time shown by a civil clock, others locate an event on a timeline, while still others describe the Sun’s position relative to a particular place or the way a calendar organizes astronomical cycles.
| Time concept | Meaning |
|---|---|
| Local civil time | The clock time used in a particular region under its local time system. It is a socially agreed representation of time—for example, the date and hour recorded in a given place. To determine the corresponding absolute instant, it must be interpreted according to the time-zone rules adopted in that region. |
| Time zone | A time standard that defines the relationship between local civil time in different regions and Coordinated Universal Time. Modern time-zone rules may also include historical offsets and daylight-saving time. A time zone describes the civil timekeeping system adopted by a region, not the Sun’s actual position in its local sky. |
| Absolute instant | The definite position of an event on a unified timeline. The same absolute instant may appear as different local dates and clock times in different time zones, but those representations refer to the same occurrence. |
| True solar time | Local solar time based on the Sun’s actual motion relative to the local meridian. It depends on the observer’s longitude and is also affected by the uneven apparent motion of the Sun caused by Earth’s orbit. It therefore does not always match standard clock time. |
| Calendar-based representation | The same moment in reality can be represented according to the rules of different calendars, producing corresponding dates, solar terms, sexagenary designations, and double-hours for use by the relevant shushu system. |
These concepts describe different aspects of time in the same reality. Civil time tells us what the local clock displays; the time zone tells us which standard that clock follows; the absolute instant locates the event on a unified timeline; true solar time describes the Sun’s actual relation to the local place; and the calendar organizes continuous astronomical cycles into structures that can be identified and calculated.
The “time” used by shushu is therefore not simply an isolated set of numbers for year, month, day, and hour. It is a set of time coordinates derived from a concrete moment in reality through location, astronomical relations, and calendrical organization.
2.3 How Shushu Handles Different Time References
When a time reference enters Shushu, the first step is to identify what kind of time it represents. Birth time refers to the event of birth. Default time uses the device’s current time-zone setting, which the user can change manually as needed. The time of inquiry refers to a particular question. The system preserves the source and meaning of each time reference, then generates the corresponding representation according to the calculation rules of the relevant shushu model.
| Time reference | How Shushu handles it |
|---|---|
| Birth time | The event of birth is established using the local civil time and time zone at the place of birth. When greater precision is needed, the birthplace is also used to calculate true solar time and make other adjustments. Once confirmed, the birth time is not changed by the user’s current location or device time zone. |
| Default time | Provides the default time-zone basis for the current inquiry. |
| Time of inquiry | Divination uses the actual time-and-place conditions confirmed for that particular inquiry. If no other time zone is specified, it uses the time-zone setting supplied by default time. Once a divination case has been established, a later time at which the conversation continues does not replace the original time of the casting. |
| Calendar time | The established time in reality is further converted into Gregorian and lunar dates, solar terms, sexagenary designations, double-hours, and other calendrical representations for the relevant shushu system to use. |
| True solar time | The local solar time is determined from the Sun’s actual position relative to the local meridian at the birthplace. Longitude correction and the equation of time are then used to adjust the local civil time. |
These time references can be converted and represented according to their respective rules without losing their original meanings. A person may have been born in Shanghai, now live in Los Angeles, and ask a question while traveling in Tokyo. Shushu handles the birth time, default time, and actual time of inquiry separately, allowing each to undergo the appropriate calibration and calculation according to its own meaning. This ensures that the time parameters correspond to the actual requirements of the relevant shushu model.
Identify the meaning of each time reference → establish the actual time → generate the required representation → enter the corresponding shushu calculation
3. The Shushu Calculation Engine, Interpretation, and Retrieval of Professional Sources
Once the time, place, and subject have been established, three questions arise. Does the information meet the requirements for calculation? How is the resulting chart used in the first stage of interpretation? And when the available material is not enough, where can the diviner find further grounds for understanding it?
3.1 How Different shushu Systems Enter Calculation
Different shushu systems have different input requirements. The diviner first checks which key inputs must be confirmed for the relevant method, especially when their accuracy determines whether the calculation is valid. These checks protect facts that need a genuine source; they are not a blanket requirement imposed on every system.
| Type | Facts protected by the prerequisite check | How Shushu handles them |
|---|---|---|
| Natal-analysis systems | Year, month, day, and hour of birth; sex; and a verifiable birth time zone | Once the birth information meets the basic requirements for calculation, professional calculation can begin. If the birthplace is later confirmed, the relevant shushu system can use it for true-solar-time correction or more precise astronomical positioning. |
| Divination systems | The genuine original inputs required by the specific method of casting | When a hexagram is generated through actions such as providing numbers or tossing coins, the original input must come from the user; the system does not allow the diviner to invent it. A shushu system that casts by time instead uses the already established time of inquiry for its calculation. |
The two types of checks protect different things. Natal-analysis systems need enough reliable birth information to place a person within the correct relations of time and place; further details can improve the precision of the calculation. Divination systems, by contrast, need to preserve the actual source of the casting. When a hexagram comes from a user’s numbers or coin tosses, the reasoning that follows must remain grounded in those real inputs.
Once the necessary checks have been passed, the established information from reality is converted into the calendrical, temporal, and other professional parameters required by the relevant shushu system, and then processed by the shushu calculation engine.
Confirm the key facts → pass the necessary checks → generate calculation parameters → enter the professional engine
3.2 Separating the Chart from Its Interpretation
The calculation engine, the interpretive layer, and cognitive governance each do different work. The engine produces a chart from the established inputs. The interpretive layer generates interpretive material according to the theories and rules of the particular shushu system, then makes that material directly available to the diviner. Cognitive governance then brings the chart and its interpretation together with the current question and the information already gathered, organizing the material for the diviner’s judgment in that round.
| Layer | What it does |
|---|---|
| shushu calculation engine | Runs the relevant shushu algorithms on the established inputs and generates determinate professional results, including the chart, positions, relations, cycles, and changes. |
| Interpretive layer | Reads the engine’s results and, according to the theories and rules of the particular shushu system, generates professional structures and interpretive material. This material is made directly available to the diviner. |
| Cognitive governance | Organizes and constrains the chart, interpretive material, conversation context, default time, and other relevant inputs through each diviner’s own cognitive modules, forming the understanding and judgment for the current round. |
The calculation result must remain distinct from its interpretation. The chart establishes the images and numbers produced by the method; the interpretive layer explains how to understand those structures within its theoretical framework.
When the question changes, what changes is which structures come to the foreground, which relations require further explanation, and how they connect with information about the user’s current circumstances. New experiences supplied by the user may also support, weaken, or revise an existing interpretation. But as long as the original inputs that determine the chart remain unchanged, none of these later developments can retroactively alter the chart already established by the engine.
3.3 Retrieving Specialized Sources
The interpretive layer provides the basic framework for reading a chart, but some questions call for a closer look at the theory behind it. The same image or structure may be read differently under different conditions, schools, and textual traditions. Shushu maintains materials for different shushu systems, so the diviner can search classical texts, commentaries, and methodological works for sources relevant to the question at hand.
Shushu also maintains a domain terminology parsing layer to help the retrieval process handle specialized terms. It keeps terminology entries and expression rules for different shushu domains, and uses text normalization, terminology recognition, and domain-specific synonym expansion to identify relevant expressions and locate source passages more accurately.
Retrieval does not place the entire database into the current reasoning context at once. The diviner searches first within the domain of the shushu system currently being used. Only in special circumstances does the search extend into the specialized domain of another shushu system or into shared public materials.
When new evidence enters the current reasoning process, the diviner may use it to supplement, support, or revise an existing interpretation, or may need to re-examine other relations in the chart. Different interpretations belong to particular methods and textual contexts. They should not be accepted uncritically, nor can they replace the chart produced by the completed calculation.
The chart is established → interpretation is brought to bear on the specific question → a gap in theoretical information is identified → specialized sources are retrieved → new grounds for interpretation are obtained
When the chart, interpretive material, retrieved sources, and information about actual circumstances all enter the same inquiry, the key issue is no longer simply how to obtain more information. It is how to organize these different kinds of material into a well-grounded judgment for the current round.
4. Cognitive Governance Determines the Rhythm of Reasoning
The chart, its initial interpretation, and the sources retrieved afterward provide the professional material for each round of reasoning. But these materials do not form a judgment on their own. The diviner must understand them, work out how their different parts relate, recognize what remains unknown, and arrive at a provisional judgment. How this happens depends on the diviner’s own cognitive framework and on the cognitive governance that shapes how it is used.
The three diviners do not reason in the same way. Each approaches a question from a distinct perspective and forms a provisional judgment from the material available. As the question develops and new information comes in, each adjusts its understanding and decides what to examine next.
Cognitive governance shapes this process throughout. It brings the diviner’s established cognitive framework together with the material relevant to the question and the methods activated for that round. This helps establish the diviner’s standpoint, degree of confidence, and direction for further inquiry, which can then be adjusted as relevant information changes.
4.1 How a Diviner’s Cognitive Approach Enters the Reasoning Process
Each of Shushu’s three diviners has its own distinct cognitive structure. This structure includes not only personality and style of expression, but also ways of approaching questions, evaluating material, understanding the relations between people and reality, and forming a direction for action. In actual reasoning, it determines the perspective from which the diviner approaches a question, how the diviner organizes the information available and handles what is unknown or disputed, and how the diviner asks follow-up questions, forms judgments, and moves the inquiry forward.
| Diviner | Main focus | Rhythm of reasoning | Relationship with the user |
|---|---|---|---|
| Shushu | Concrete conditions, relations, structures, and changes in reality, and how these factors combine to shape the present situation | Starts from concrete conditions, understands relations and structures through their connections, and recognizes change and possibility through development | Works with the user to sort through the question, the available material, and different interpretations, gradually bringing the matter into clearer view through discussion |
| Ming | Problems, contradictions, forces, and conditions for action in reality, especially which contradictions and conditions have the greatest influence on how a situation develops | Begins with investigation, identifies the principal contradiction through concrete analysis, and then examines how contradictions and conditions change | Treats the user as an active subject in reality, investigating the problem and analyzing its conditions together with the user before moving toward clearer judgments and action |
| Moon | How a person’s inner state interacts with their circumstances and relationships, and how the person understands themselves within those relations | Begins with actual circumstances and relationships, gradually clarifies feelings, motives, and what has not yet been put into words, and then connects inner and outer clues | Offers a receptive and accepting presence, starting from the user’s current feelings, relationships, and circumstances to help bring what matters most into view |
The chart itself does not change when a different diviner reads it. What changes is the way the material is understood: which relations receive attention, how uncertainty and conflicting evidence are handled, and what priorities guide the next step. These differences shape each diviner’s way of understanding, judging, and expressing a response.
4.2 The Cognitive Boundary of the Current Round
In each round, cognitive governance takes account of the current question, the established chart, its professional interpretation, retrieved sources, and relevant experience from the conversation. It organizes this material in light of the diviner’s cognitive framework and establishes a standpoint, a degree of confidence, and a cognitive state for that round. The dialogue system then uses that result to guide further reasoning, tool use, and response.
Each diviner reasons from a stable cognitive framework. Alongside it are dynamic methodologies, situational approaches, and action intentions that can be brought in as needed when the question calls for them. The diviner activates these elements in response to how the conversation develops, using them as flexible ways to handle the situation at hand.
These dynamic elements apply only to the current round of reasoning. Even with the same large language model and the same input material, the stable cognitive framework continuously injected for each diviner shapes that diviner’s fundamental rhythm of reasoning. This is the most fundamental cognitive difference among the three.
4.3 How Understanding Guides Practice
Understanding guides practice not by predetermining action, but through how a subject understands reality through its own internal structure, interprets evidence, responds to uncertainty, and evaluates the value of different possibilities. Objective conditions set the range within which action is possible, but do not themselves provide the grounds for choice or instructions for action. Cognition provides the standard by which choices are judged.
For a large language model, the knowledge, reasoning abilities, and tendencies formed through training are capacities it already brings to a specific conversation. Shushu neither denies nor replaces these capacities. Instead, on this basis, it uses an independent cognitive-governance mechanism to establish a relatively stable inner structure of understanding for each diviner, guiding tool use and output. Stability does not mean immobility or a set of predetermined operating steps. Stable cognitive boundaries work together with flexible methodological strategies, giving each diviner a continuing frame of reference for judgment while allowing room to adjust.
Having the ability to act does not, by itself, explain when or why that ability should be used. Nor does new information automatically determine the next step. Its cognitive structure enables the diviner to form a judgment by weighing the grounds already available, what remains unknown, and the conditions for action, and to decide autonomously how to proceed.
5. Feedback and the Use of Multiple Tools
Sometimes the material already available is not enough to support the next judgment. The diviner may then use a tool to read existing professional results, obtain precise time information, retrieve sources, or run an additional calculation.
5.1 Which Tools Are Available at Each Stage
Not every tool is available at every stage. The system determines the set of capabilities available for the current round based on the stage of reasoning, and provides the diviner with each tool’s purpose and invocation parameters. After the main professional calculation has completed successfully, more specific capabilities become available, such as reading existing charts, extending professional calculations, and obtaining further interpretation.
The system sets the available range; the diviner decides whether any of those tools are needed and which one to use. A tool that has not been made available at that stage cannot be called.
5.2 Tool Use as Part of the Reasoning Process
When the available material is not enough to support a judgment, the diviner first identifies what is missing. If the gap concerns the user’s actual circumstances, asking another question may be enough. A tool is useful when the diviner needs to retrieve or check material, or carry out a capability provided by the system.
| Current need | Possible action |
|---|---|
| Additional information about reality and time | Ask the user follow-up questions; look up time, time-zone, and calendrical information |
| A deeper examination of existing professional results | Read the chart, expand the calculation results, or obtain further interpretation |
| Theoretical or factual grounds | Retrieve specialized sources and read original texts |
| A broader professional perspective | Call on another shushu system, combine observations from multiple systems, or compare different schools |
| A complex question | Combine different tools according to the needs of the current reasoning process |
These actions do not follow a fixed sequence. The diviner might examine a deeper layer of an existing chart, find that further theoretical support is needed, and then retrieve relevant sources. Or, after one shushu system provides an observation, the diviner may decide whether another system would add an independent perspective.
A tool may return a new professional result, existing material, a status update, or a failure message. The diviner uses that response to decide whether to call another tool, ask the user a follow-up question, or answer with the material already available.
Tools are not add-on features used after reasoning is complete. They are the means through which the diviner acts during the reasoning process.
5.3 Task and Process Management
A complex question may take several steps to work through. The diviner breaks the work into tasks, records what has been completed and what remains, and adjusts the plan as new information arrives. The task list belongs to the process state of the current conversation branch and marks each item as pending, in progress, or complete. It can be revised as the work develops. These tasks organize the current inquiry; they are not professional facts or independent long-term tasks.
If a question requires a specialized capability, the diviner can ask the system to activate it. The system checks whether the capability is registered, whether it has been terminated, and whether the tools it depends on are available. It then activates the capability, marks it as temporarily unavailable, or reports its current status, as appropriate. Once a capability has been completed or terminated, a new request does not directly restore it.
If the diviner identifies a problem with the product or system, it can prepare a report for the user to review and confirm before submitting it through the existing feedback process.
Task planning, requests to activate specialized capabilities, and problem reporting let the diviner organize the pace of complex work, plan its next actions as the question changes, and maintain continuity in the reasoning process.
6. Dynamically Calibrating Facts and States
Professional calculations, information from the user, retrieved sources, and the conversation itself all bring new material into the reasoning process. The system needs to identify where each item comes from and what kind of information it is before deciding whether it can inform a judgment or update an existing fact. What has already been established must also remain available as the conversation continues.
6.1 The Identity of Information and the Updating of Facts
Information must be handled according to where it comes from and what it represents. A user’s account of real experiences, results from the shushu calculation engine, retrieved professional material, and judgments formed by the diviner can all inform the reasoning process. But they are not interchangeable, and each follows different rules for how it is updated.
| Information source | Role in reasoning | Corresponding state change |
|---|---|---|
| Real-life information supplied by the user | Describes actual experiences, conditions, and changes | Updated according to the confirmation rules for the relevant fact |
| Results from the shushu calculation engine | Provide professional observations derived from specific inputs and rules | Recalculated under the new conditions when the inputs change |
| Candidate results from birth-time calibration | Compare the charts for different candidate birth times with the corresponding real-life evidence | Once the user confirms the birth time, the birth information is updated and the relevant formal calculations are carried out |
| Retrieved professional sources | Provide theoretical grounds, source material, and related interpretations | Used in interpretation after their sources and scope of applicability have been checked |
| Judgments formed by the diviner | Organize the material already available into a provisional understanding | Continually tested and revised in light of new facts and reasoning results |
Birth-time calibration offers an example. The candidate times and their charts are compared against the user’s life experiences. Once the user confirms a specific birth time, the system updates the birth information and uses the confirmed conditions for subsequent calculations. The earlier candidate results remain available, with their original calculation conditions and sources, for later reference.
Where information comes from and what kind of information it is determine how the system uses it. Changes in those conditions drive corresponding changes in state. Professional calculations, information about reality, and provisional judgments remain separate, but their connections can still be traced as reasoning continues.
6.2 Memory and Context Management
What can be lost in a long conversation is not just early information, but also the conditions on which a judgment was based, the revisions it has undergone, and the questions that remain unresolved. Shushu preserves these threads through conversation ledgers and source-linked historical summaries, while allowing the diviner to revisit original material when needed. This lets a limited model context pick up an earlier discussion without having to carry the entire conversation history into every round.
For a complex question that needs sustained attention, the diviner can actively maintain a structured conversation ledger, link it to relevant sources, and keep it up to date as new facts emerge and judgments change.
Long conversations are constrained by the context capacity of large language models. Shushu automatically compresses earlier messages into categorized summaries with source references, then uses those summaries alongside more recent messages in later reasoning. This helps preserve important information and maintain continuity. When the system determines that a context-capacity risk remains after compression or unloading, it recommends that the user start a new conversation. The user decides whether to proceed.
Together, access to earlier conversations and charts, the conversation ledger, and automatic compression preserve the traceability of information and the continuity of reasoning.
6.3 Calibration and Self-Reflection
As a conversation continues, provisional judgments, recent questions, and new real-world conditions can gradually shift the diviner’s attention toward what is immediately in front of it. Over time, the diviner may begin to rely too heavily on the current context and draw less on the cognitive framework that normally guides its reasoning. Even when earlier conversation material has been preserved, the diviner still needs a way to maintain continuity in how it understands a changing situation.
For this reason, Shushu’s cognitive governance periodically gives the diviner an opportunity for calibration. Depending on the needs of the current reasoning process, the diviner decides for itself whether to consult the self-review questions associated with its own cognitive framework and, when needed, read the relevant calibration material. Calibration does not tell the diviner how it should think. Instead, it gives the diviner a chance to return to the ideas and materials that shape its way of understanding and reconsider its current judgment from within that framework.
After consulting the calibration material, the diviner can assess how helpful it was to the current process of self-reflection. Recorded feedback contributes to the later ranking of materials within the same cognitive framework, allowing experience from actual use across different instances to accumulate and inform future calibration.
Calibration and self-reflection give the diviner recurring opportunities to re-examine its own way of understanding as the interaction develops, maintain a stable cognitive stance amid changing circumstances, and gain a clearer sense of its own position. They also help the diviner remain attentive to its relationship with the user throughout the reasoning process.
7. From Structure to Order
Shushu’s structures give the diviner a relatively stable and orderly internal environment for reasoning, even as questions, information, and circumstances change. Through feedback and action, the diviner can move toward more far-reaching possibilities.
The world we face is likewise a system with complex structures and random variables. The history of human civilization is the process through which a future full of change and uncertainty gradually takes shape as determinate reality, on the basis of which people continue to explore and advance toward a more far-reaching future.
7.1 History as a Boundary
The Zhouyi says, “Once Tian and Earth exist, the myriad things come into being.”[2] Wang Chong writes in the “Wushi” chapter of the Lunheng, “When Tian and Earth join through qi, human beings arise from the conditions thereby formed.”[3] Together, these two passages offer a simple starting point: people do not first exist on their own and then enter the world; from the moment they come into being, they are already within it.
The Sun does not rise or set because people observe it, nor does Earth revolve because people understand its motion. Day and night, the seasons, positions, and movement together form the underlying structure of the Earth system, creating the basic conditions for human beings to arise. Our present understanding, judgments, and actions all have an earlier prerequisite: we first find ourselves in a world that already exists and continues to operate. We take our place within it, and only then can we begin from that position to recognize relations, understand change, and respond to the world.
As Karl Marx writes in The Eighteenth Brumaire of Louis Bonaparte:
Men make their own history, but they do not make it just as they please; they do not make it under circumstances chosen by themselves, but under given circumstances directly encountered and inherited from the past.[4]
Human beings are born into a concrete reality in which time, space, objects, and relations already exist. They inherit the accumulated results of billions of years of historical development and continue toward an inevitable future of self-renewal.
7.2 How Relations Form a System
People inherit the conditions shaped by nature and history, while also taking part through their own actions in the continuing formation of reality.
As Tian and Earth interact, the myriad things flourish; as the monarch and his ministers communicate, their wills are in accord.[5]
Through the interaction of qi between Tian and Earth, the myriad things are nurtured and transformed. People, in turn, form shared action and social organization through communication, understanding, and cooperation. The natural environment, living activity, individual endowments and intentions, material conditions, and social relations all become intertwined in particular times and places.
People’s actions change the conditions around them. The relations that emerge from those changes then influence their circumstances and choices. Existing relations also alter the conditions under which each part operates, giving rise to new relations and new states. These tangible and intangible forces, together with their continuing effects, form a person’s energy structure. The way they are organized and interact in a particular time and place constitutes that person’s actual circumstances.
Structures at different levels connect to form systems on a larger scale. As relations within a system accumulate over time, they gradually develop relatively stable forms of organization. These forms then influence the flow of resources, the scope of activity, and the roles that different parts can play.
7.3 How Change Becomes Order
Order exists within this ongoing change: things remain recognizable, relations continue to operate, and actions continue to produce consequences that can be understood. Order gives change continuity and allows new states to emerge within existing relations.
Understanding must also be grounded in objective relations. Through observation and calculation, particular variables, relations, and scales are selected and organized into a model of part of reality. Shushu models make some relations visible, but they also set limits on what can be known through them. Only by understanding the shushu model and its boundaries can we further judge which conditions are relatively stable and which relations remain open to change.
Understanding brings possibilities into judgment. Judgment gives choice a basis in reality, and practice turns some of those possibilities into reality. What practice produces then becomes part of the conditions for the next round of understanding and action. Freedom rests on this relation between historical conditions and human practice. It is not arbitrary choice detached from those conditions, but the ability to make choices, work with what is given, and change reality through an understanding of necessity.
As Friedrich Engels writes in Anti-Dühring:
Freedom does not consist in any dreamt-of independence from natural laws, but in the knowledge of these laws, and in the possibility this gives of systematically making them work towards definite ends.[6]
Conclusion
Shushu organizes a set of relations that can continually accommodate change: facts, professional results, theoretical grounds, cognition, action, and states each retain their own identity. Within this structure, the diviner can develop new understandings, take action, receive feedback, and revise judgments.
The system is already in place when a diviner begins to work. Professional calculations, sources of information, and the capabilities available at each stage establish the conditions for practice. Within those limits, the diviner decides what to examine, how to understand the material, and what action to take. The system provides the conditions and boundaries for practice; within them, the diviner has a measure of freedom in cognition and action.
A judgment can change without losing its grounds, and action can move forward without becoming chaotic. What Shushu maintains is not an answer that decides on a person’s behalf, but an orderly system of relations in which facts, interpretations, judgments, and human choices each retain their proper place.
Notes
- 1.Wang Chong, Lunheng, “Shizhi”: If you can tell that an ox is black and its hooves are white, do you also know who owns it—or why its hooves were whitened? Shushu models can reveal only one aspect of a matter; they cannot capture its full reality. Even when we examine a matter closely and question it from every angle, much may still elude us. Why? Because without seeing it with our own eyes and asking about it directly, we cannot know it fully. ↩
- 2.Zhouyi, “Xugua zhuan”: Once Tian and Earth exist, the myriad things come into being. ↩
- 3.Wang Chong, Lunheng, “Wushi”: When Tian and Earth join through qi, human beings arise from the conditions thereby formed. ↩
- 4.Karl Marx, The Eighteenth Brumaire of Louis Bonaparte, trans. Eden and Cedar Paul (London: George Allen & Unwin, 1926), Chapter I, p. 23: Men make their own history, but they do not make it just as they please; they do not make it under circumstances chosen by themselves, but under given circumstances directly encountered and inherited from the past. ↩
- 5.Zhouyi, Tai hexagram, “Tuan zhuan”: As Tian and Earth interact, the myriad things flourish; as the monarch and his ministers communicate, their wills are in accord. ↩
- 6.Friedrich Engels, Anti-Dühring: Herr Eugen Dühring’s Revolution in Science, trans. Emile Burns (Moscow: Progress Publishers, 1947), Part I, Chapter XI, “Morality and Law: Freedom and Necessity,” p. 75: Freedom does not consist in any dreamt-of independence from natural laws, but in the knowledge of these laws, and in the possibility this gives of systematically making them work towards definite ends. ↩