Tomorrow There Will Probably Be Orders
On Thursday morning, Gu Ning stuck a note on the edge of the table at the Chengwan Repair Station that read: "Tomorrow there will probably be orders." Tang Ke glanced at it and said the client had asked about table-and-chair repair the day before, in a friendly tone. Gu Ning was about to write down her judgment in the log, only to discover she did not know at what later moment she should mark it as correct.
The Chengwan Repair Station, Gu Ning, Tang Ke, and the temporary assistants appearing in this chapter are all fictional constructs of this book. We begin with a repair commission still awaiting confirmation and gradually add predictions, messages, and outcomes, without treating their experiences as the author's firsthand accounts or real business data. Here there is first of all no complete statistical table, only a familiar word of approximation.
Tang Ke held that if the client replied and was willing to keep talking, that already counted as an order. Gu Ning held that specific scope had to be confirmed. Another temporary assistant, on hearing the word "order," thought of the other party as having already paid. The three appeared to endorse the same prediction, yet in fact were waiting for three different outcomes.
If the other party kept asking questions the next afternoon, Tang Ke would say she had guessed right, Gu Ning would say it had not yet begun, and the assistant would say the money had not arrived. The dispute was not because anyone could not calculate; it was because the object of calculation had not yet been formed. A seemingly shared word preserved different exits for different observations.
The task of this chapter is to first make that approximation into a specific question that can be judged. We will not rush to fill in a percentage. Without a clear event, the more precise the number, the more likely it packages ambiguity as reliability; even if the guess ultimately proves right, one does not know what was guessed.
To What Does Probability First Attach?
Everyday language attaches possibility to many objects. Tomorrow may be busy, the table and chairs may need repair, the client may agree, the repair station may earn more. All of these are meaningful, yet they do not belong to the same event. Business may occur while the commission remains unconfirmed; confirmation may occur while the income has not yet arrived.
This book understands an event as: within an already specified observational range, whether a set of conditions holds. It is not a summary of all of reality, but an outcome object identified out of a continuous process. To judge, one needs to know which changes would make it hold and which similar changes still fail the conditions.
Gu Ning's note wrote only "orders," without saying which commission, nor what "orders" referred to. If tomorrow someone happened to bring another small stool, she could also claim she had guessed right. The object of prediction would then shift with the outcome, and any later favorable fragment could be gathered back into the original statement.
A probability judgment needs first to attach to a relatively stable question. Stability here does not mean the object never changes, but that when this prediction is checked, the question cannot silently become another matter. New questions may arise in the future, and the original question must still preserve the range it had at the time.
This requirement is not only to prevent lying. Many people sincerely believe they have been talking about the same object, only to change the meaning of a word once news arrives. Preserving the event conditions gives people a chance to discover such changes, rather than maintaining judgment by arguing over whose memory is more accurate.
Distinguishing a Commission from Many Needs
Tang Ke located the commission under discussion and gave it an identifier: the seventeenth request in the repair station's log, abbreviated R17. It concerned a batch of tables and chairs already listed in an inventory, excluding another possible addition of cabinets and excluding inquiries from other clients. The identifier merely makes it convenient to return to the material; it confers no additional authority on the request.
An object identifier can reduce confusion, but it cannot complete the definition by itself. If the number of tables and chairs, the repair scope, and the receipt period under the same identifier keep changing, what is confirmed later may no longer be the task originally discussed. Gu Ning needs to know which version she is predicting, not merely to remember a number.
So Tang Ke saved the current description and called it the first version. It listed the object under discussion, the proposed repair scope, the payment conditions, and the receipt period. The first version had not yet obtained the client's confirmation, but it at least let both sides know what was to be confirmed. Later proposals could be saved separately without overwriting this raw material.
A version is not a demand that life stop changing. The client may discover omissions; the repair station may adjust arrangements. Keeping versions is meant to present changes as changes, rather than letting an old prediction automatically own all the later new conditions once the outcome appears.
Gu Ning can now ask a smaller question: whether the client will confirm this first-version request. She still does not know the answer, but she now knows what the answer is about. Shrinking the object does not reduce the difficulty in reality; what it reduces is the part of the judgment that can be substituted at will.
What Outcome Does Confirmation Require?
Tang Ke said the other party had already expressed willingness to consider. Gu Ning asked whether willingness to consider equals acceptance of the current scope. Tang Ke admitted it did not, because the other party might still ask to reduce the quantity, change the period, or keep asking about payment. A positive attitude can become material for later judgment, but it cannot directly serve as the outcome that establishes the event.
For this exercise, the two agreed that confirmation means receiving the client's explicit acceptance of the first version's task scope, payment conditions, and receipt period. This condition belongs to a thought experiment and does not presume to be the general standard for how real contracts come into being. It only gives this book's prediction a commonly checkable object.
Confirmation also does not equal the other party having paid, still less the repair having been completed. An event's holding can be a condition for subsequent activity, but it does not bring all subsequent outcomes with it. If Gu Ning later wants to predict whether payment occurs or whether completion happens on time, she must set up a separate question.
Such decomposition frees judgment from bearing the entire story. After commission confirmation, things may still change; during fabrication, deviations may still occur. The original event is not proven never to have happened because of later difficulties, nor does later success retroactively prove that the initial confirmation must have arrived.
Separating request, acceptance, payment, and completion protects the role of each piece of material. If all were packed into the word "order," Gu Ning might use a single inquiry to prove income was available, and use final payment to prove she had long known production would go smoothly. The further the scope expands, the harder it is for evidence to correspond.
Until When, and Why It Must Be Written Down
The original note said "tomorrow," without specifying a cutoff point. With no reply in the morning, Gu Ning could wait until afternoon; with no reply in the afternoon, she could say evening still counts as tomorrow. If the message the day after were then called "a bit later," the original judgment would lose any place at which it could end.
The observation window for this prediction is provisionally set to Friday 17:00, inclusive of the 17:00 boundary. This point in time is a convention within the case, not a general rule of the table-and-chair repair trade. It turns "whether confirmation occurs" into a question with a window, and also gives later confirmations a place to enter a new record.
The window affects the meaning of the outcome. No confirmation before Friday 17:00 does not mean the other party will never commission; a confirmation received over the weekend also cannot automatically rewrite the prediction in the original window as fulfilled. Both can be accurate at once: the original event did not hold within the specified window, and a new confirmation later occurred.
How exactly to freeze the prediction time point and how to handle messages learned only after the deadline will be developed in the next chapter. This chapter first keeps the window, without stuffing every detail of occurrence time and message-discovery time into the definition at once. Each added condition should address a problem it actually raises.
A cutoff point is not a command to the future. It does not oblige the client to answer on time, nor does it prove that a shorter window is more reasonable. It merely lets this question know when it accepts checking, preventing judgment from hiding indefinitely in the not-yet-finished future.
What Material Can Serve as Outcome Evidence
The two can stipulate that an explicit acceptance sent by the client through the agreed communication channel is the primary material for this check. Gu Ning saves the original message, not merely Tang Ke's note saying it was "pretty much settled." If the other party says only "received," it does not satisfy the confirmation condition above.
The material condition does not mean the message form is always unique. If the client explicitly accepts in person, a checkable record may also be obtained. But one should specify in advance how it will be preserved, who observed what, and whether such material suffices to identify the same acceptance object. One cannot wait to see a favorable form before lowering the requirement on the spot.
Outcome evidence also differs from prediction evidence. The other party's earlier detailed inquiries can influence Gu Ning's judgment about the future; they do not directly become material that a future confirmation has occurred. The former is used to estimate an outcome not yet arrived; the latter is used to check whether an outcome holds.
If the message content remains ambiguous, a pending-check state can be retained. One cannot, because Gu Ning needs to fill a cell, compress "may accept" into "has accepted"; nor, because it cannot be determined immediately, write it as an explicit refusal. Unknowns about material need a place of their own.
Here a boundary of the record already appears: the event can be defined clearly, while observation may still be incomplete. A clear definition is not the acquisition of omniscience; precisely the opposite, it enables one to point out which piece of material is missing. Without a definition, even what one does not know is hard to state clearly.
Why Similar Outcomes Do Not Automatically Count as the Same Outcome
First set a comparison case: at noon on Friday, the client is willing to repair half of the tables and chairs and requests a different receipt period. This may be a valuable new collaboration, but it has not accepted all the conditions of the saved first version. If the original prediction concerned only the first version, it cannot directly be recorded as that event holding.
Gu Ning may set up a broader question instead, predicting whether any acceptable repair commission forms. It might be closer to Tang Ke's actual need, but it must then specify which changes still belong to this object and which have crossed the line. A broad question is not inherently wrong; it is a broad question without boundaries that lets the verdict drift with the outcome.
Set another comparison: the other party fully accepts the first version, then at a later point proposes cancellation. Whether the original confirmation occurred and whether the later cooperation persists must be separated. Gu Ning cannot, because the cancellation is disappointing, declare that confirmation never occurred, nor use the initial confirmation to prove that payment would necessarily follow.
Different questions can be retained simultaneously, but they cannot borrow answers from one another. A narrower event helps check specific material; a broader event helps observe a class of outcomes. Their probabilities and outcomes may differ, and subsequent samples may not be directly mergeable.
The purpose of these comparisons is to let boundaries be tested before outcomes appear. If a rule seems clear only in smooth examples but falls into confusion upon partial acceptance and modification, the definition is not yet usable enough. Handling boundaries in advance reduces not change, but after-the-fact arbitrary interpretation.
Prediction Is Not Another Way of Writing a Wish
Gu Ning hopes the commission will hold, because the repair station needs to arrange its next stretch of work. This hope can be a genuine motive, but it cannot supply the answer to the event definition. The more one needs an outcome, the easier it is to read a positive inquiry as acceptance and a delayed non-reply as careful consideration.
Tang Ke, too, may hope things will not be settled too soon, because the equipment periods are still being adjusted. The same message will present different meanings under different needs. Separating wish from prediction does not demand that people have no emotions, but prevents emotions from quietly altering the conditions of holding.
A prediction record may separately note why one cares about the question, but the reason for caring cannot substitute for material supporting that it occurs. The repair station's great need for orders does not prove the client is more likely to confirm; Gu Ning's fear of losing the opportunity does not prove the probability of confirmation has fallen.
Within RC, concepts in turn shape observation. Applied here, this means a person may seek out messages along the interpretation they expect and ignore conditions that do not fit the expectation. Such domain mechanisms require specific material; one cannot judge a person's judgment to have failed merely because she hopes for success.
The event definition gives this feedback a point of return. When new material arrives, first ask whether it satisfies the written conditions, then ask how it changes the future judgment. Wishes may still exist, but they no longer alone decide what counts as an outcome.
A Prediction Holding Does Not Decide Our Action for Us
If Gu Ning later thinks confirmation is quite likely, whether Tang Ke should immediately free up all equipment time remains another question. Preparation needs support and may also affect other tasks. Prediction discusses the tendency of outcomes; action discusses what can be borne under certain outcomes. The two cannot be directly merged.
This volume first addresses the formation and calibration of probability judgment and does not, in Chapter 1, issue an investment plan for the repair station. Even with a complete event definition, a prediction may be wrong; even with a fairly reliable judgment, action may still require a different arrangement because losses are irreversible. The next volume will continue from these separated conditions.
Likewise, an event holding is not a moral verdict. The client's failure to accept within the window does not directly prove he is untrustworthy; Tang Ke's somewhat optimistic judgment does not directly prove she intended to mislead. Assessing causes and responsibility requires separate material about commitments, information, and actions.
Keeping judgment, action, and evaluation separately preserves a single outcome from bearing too much meaning. Gu Ning needs to know how well she predicted, without explaining everyone's character through one outcome or scoring the original judgment by whether the cooperation ultimately went well or badly.
Probability representation is therefore not the master switch of life. It can help a certain class of questions be compared more clearly; it cannot alone cover all needs, values, and commitments. Boundaries do not weaken the tool; they prevent the tool from acquiring unlimited authority where there is no material.
Defining Something Together Is Not Creating the Outcome Together
Gu Ning and Tang Ke agreeing on what counts as confirmation is a preparation for joint observation. It lets both parties later return to the same material; it does not oblige the client to accept merely because the two agreed. The definition changes the object of judgment; it does not thereby create tables, replies, or payments.
RC understands cognition as a gradual and incomplete mapping of a Ground of Possibility, with judgment marking the current boundary of cognition. From this starting point, this book treats an event as a representation of an outcome within limited observation. It is neither a transparent copy of all reality nor a fiction that people can command at will.
The domain bridge lies in: participants choose objects and conditions and preserve relevant observations, so that later outcomes can be checked against the same range. This process requires language, records, and actual material. Conceptual similarity alone cannot substitute for a confirmation message having actually appeared.
Hence Gu Ning cannot say that since the order is constructed, merely renaming it makes the prediction correct. The limited role of construction comes from jointly maintaining conditions, not from making every interpretation equally reliable. Changing conditions without explanation destroys precisely the path by which this judgment could originally be returned.
Likewise, a clear definition does not mean future probability has been determined. We have merely given the subsequent collection of samples, comparison of conditions, and judging of evidence an object. How a mathematical value acquires support still requires new steps and cannot be generated automatically by clear language.
How to Preserve Objections to the Definition Itself
Tang Ke proposed that the first-version confirmation was too narrow. What she really wanted to know was whether the repair station would obtain work it could schedule, not whether the client accepted the original description item by item. This objection is not an attempt to blur the rules, but a point that the definition may not address the actual concern.
Gu Ning may adjust the question before the prediction is issued and write out the new range. She may also keep both questions, explaining what each answers. Which definition to choose depends on use and observable conditions, and it should not be preferred merely because it is easier to guess right.
If the prediction has already been issued and the object later proves inappropriate, the old question should be retained, a revised version set up separately, and the reason for the change stated. The new version may be more useful, but it cannot borrow later material to claim an earlier issue time than the old prediction.
A definition may also carry too high an observation cost. If the client is required to reconfirm every detail repeatedly, the labor needed to obtain outcome material may exceed the current use. Irrelevant fields can be reduced while the key conditions deciding whether the event holds are kept; precision is not infinite subdivision.
Processual completeness here manifests first as: the definition can be pointed out as deficient, and it can be revised according to use, without erasing the original question. Stability makes a judgment reviewable; revision lets judgment continue into real needs. The two capacities must coexist.
Not Holding Does Not Necessarily Equal Explicit Refusal
Reduced to whether confirmation occurs, the event seems to have only two cells. But the actual process can still have many states: no reply, a reply without acceptance, a proposed modification, an explicit refusal, and full acceptance obtained. Whether each holds under this condition can be checked separately, and their causes cannot all be handled by a single "not holding" label.
Suppose the other party proposes modifications on time: this material shows the old version was not fully accepted, while possibly providing support for a new version. If Gu Ning writes all instances of not holding as refusals, later analysis of conditions will merge people who need negotiation with those unwilling to continue into a single group. Simplifying the outcome does not save explanation; it may instead create a wrong explanation in advance.
Therefore, the necessary original state can be preserved beside the event outcome. Whether it holds answers this question; the state description identifies why the conditions were not satisfied. The two need not contend for a single correct place. When later comparisons of refusal and modification are needed, the record will not be reduced to a bare "no" with no origin.
There is another case: Gu Ning has not obtained enough observational material and simply cannot judge whether the event holds. This differs from having observed that the conditions are not satisfied. A complete "no" indicates that checking within the relevant range has been completed; temporary unknown indicates that checking still lacks material. Counting the unknown directly as "no" would let a gap in the record quietly alter the number of outcomes.
This distinction does not require building an infinite state system for every prediction. One need only keep the important differences that affect the current interpretation; other irrelevant details need not all enter. If a new use appears later, the range can be extended, but details not preserved initially must not be written from memory as facts known at the time.
The conditions for holding and not holding should also connect. One cannot require explicit acceptance for an affirmative outcome while resting a negative outcome merely on one's own failure to see a message. Checking both sides against the same range is what prevents a very strict definition from coexisting with very loose failure statistics. The next chapter's distinction between time and observation will further address such gaps.
Writing This One Question as an Event Card
In the end Gu Ning did not write down a confident number, but saved an event card. The object is the first version of request R17; the outcome condition is the client's explicit acceptance of the saved task scope, payment conditions, and receipt period; the window runs to Friday 17:00; the outcome evidence is returnable explicit confirming material.
The card also notes that continued inquiry, partial modification, actual payment, and completed repair are not to be merged into the same outcome. If observational material is missing, a pending-check state is retained. If the definition is later adjusted, a version is saved separately without overwriting the original question already issued.
This card is not a universal template. Applied to other objects, one must anew judge what outcome is observable, which conditions matter, and whether the window is reasonable. Reusing fields can reduce repeated preparation, but cannot let the repair-commission definition answer other questions directly.
Readers can subject a prediction they often make to the same check: what object the words point to, which changes make it hold, when it is checked, what material serves as evidence, and which similar outcomes must not be counted in directly. If these questions remain unanswered, deal with the object first, and do not rush to argue over whether some probability is too high.
The judgment obtained in this chapter is that the preparation for a reliable prediction begins with an identifiable event, not with a pretty number. The event card lets us know what is to be judged and lets a later discovery of error know where the error points. It does not guarantee Gu Ning will guess right, but it gives right and wrong guesses an interpretable range.
The Chengwan main line remains at Thursday morning; the commission is not yet confirmed, and Gu Ning has not yet given a material-supported numeric estimate. The next chapter continues from here: why the same event predicted at different time points faces different information; and why an outcome occurring and an outcome being seen need to be preserved separately. The object has taken preliminary shape; time will bring the record into a genuinely checkable process.