Chapter 26. Model C, Complexity-Based Conflation, and Explanatory Calibration
I. Class Entities Do Not Exist; Class Theory Belongs to Model C from the Outset
What this book denies is neither concrete differences such as wages, labor, property, debt, housing, risk, domination, responsibility, and available exits, nor statistical classifications made for particular purposes. What it denies is compressing, projecting, and naming these differences, then reifying the resulting classification into a class that exists prior to concrete relations. Such a class does not constitute a natural community; it does not automatically possess unified interests, a unified consciousness, unified responsibility, a unified capacity for action, or a historical mission. Classifications can exist, and similar economic pressures can exist. The class entity whose existence this book denies does not exist.
The class image must first be distinguished from class theory. A class image is formed when high-weight differentiating units such as labor, income, property, debt, housing, risk, available exits, and capacity for domination undergo low-dimensional compression, projection, and naming along axes of economic, labor, property, and domination relations. The real differentiating units involved provide the material basis; the processes of extracting them, assigning them weight, connecting them, and compressing them form partial explanatory power. The class image is the low-dimensional output of this process. It can present some of the circumstances extracted, but it is not society itself, still less a subject that judges and acts on its own.
The class image is readily mistaken for an entity precisely because class theory and its interpreters compress multiple, not entirely identical relations into a recognizable name. The name allows people to recognize certain shared circumstances quickly, but it also occludes how differences are preserved and recast at different nodes. When two people with similar incomes face different consequences because of household registration, medical history, caregiving responsibilities, or local recognition, the image cannot replace tracing these differences. The image's ease of recognition and dissemination comes from compression; it does not exist prior to compression.
This distinction also preserves the proper target of criticism. Criticizing class theory does not mean denying unpaid wages, housing pressures, debt, or relations of domination. Those differences can force any explanation to adjust. What the criticism rejects is compressing a set of differences capable of producing consequences into the name of an entity, then using that name to replace inquiry into other differences, nodes, and relational chains. Conflating the two makes it easy to misrepresent criticism of the total image as denial of real harms.
The structural appeal of Model C comes from the ability of interpreters who adopt it to extract materials from real suffering. Unpaid wages, debt, and housing instability force people to seek quick, certain answers; interpreters who adopt a total image may use this to name harms, connect materials, and divide friends from enemies. This appeal does not prove the total image correct. Critics must still face the consequences it captures. If they ignore those consequences, their own arguments against the total image will also lose their grounding in real materials. The most dangerous feature of Model C is not fabrication of materials out of nothing, but the ability of interpreters who adopt it to grasp real suffering and then compress that suffering into a total image that refuses revision. Overdue wages, approaching rent payments, debts falling due, and unstable labor relations do indeed simultaneously change a family's capacity to wait, room for negotiation, and residential stability. These consequences are not illusions. Precisely because they are weighty, a closed explanation can readily appropriate them as evidence.
An entirely ineffective account is unlikely to develop into an illusory explanatory construction. It has no materials to retrieve and can hardly demand that people act according to it. Model C's power of closure comes from interpreters who adopt it first extracting a part of reality and then turning that part into the sole entry point. Thereafter, differences it has not extracted cease to be regarded as materials requiring renewed comparison. Instead, they are required to prove that they ultimately still belong within the total image.
Class theory is a typical instance of Model C. Interpreters who adopt class theory can extract real differentiating units such as labor, wages, property, housing, debt, risk, available exits, and domination, and can identify their high weights in certain relational fields. The reality and real-world efficacy of these concrete differentiating units neither prove the existence of class entities nor give class theory the standing of Model A, Model B, or an ordinary limited tool. Class theory's basic structure is Model C from the outset. It does not first exist as an ordinary limited model and become Model C only when overused. Its ability to extract a set of weighty differentiating units does not change its basic structure: from the outset, it expands distinction extraction within a limited scope into a total image, compresses the complex differentiation network into a total class image, prearranges the explanatory positions of other differences, and uses this image to monopolize explanation.
Extraction, weight assignment, compression, naming, ontologization, and demotion are operations performed by class theory and the interpreters who adopt it. Class theory presents this image as society's underlying structure and uses it to decide what is fundamental, what is superficial, and what is merely occlusion or displacement. The class image is the result of these operations. This cannot be reversed into a claim that the image itself sees differences or itself expands them into what society fundamentally is.
Thus, within class theory's closed account, nation, family, locality, religion, culture, historical indices, institutional eligibility, medical history, accent, kinship responsibilities, and community recognition are assigned positions as occlusions, appendages, displacements, or surface layers of class relations. Interpreters who adopt this theory no longer begin by asking: Do these differences enter into consequences through different nodes? Are they gaining weight at this moment? They hand the answer to the total class image in advance. This is precisely total-image formation; it cannot count as ordinary limited explanation.
II. Model C's Ontologizing Closure
Limited explanation necessarily extracts distinctions and necessarily downweights certain differences. It need not pretend to see everything, but it must state what it has extracted, what it has temporarily downweighted, and which counterexamples would force it to revise. Compression is not itself an error. Dangerous compression occurs when compression forgets what it has downweighted and refuses to let counterexamples rearrange weights, nodes, and relational chains. Model C is the extreme form of this danger: it erases the boundaries of limited distinction extraction and disguises the image's boundaries as those of reality.
Explanations that adopt Model C refuse this responsibility. Interpreters who adopt it present the results of their distinction extraction as reality's underlying structure and a crude projection as what society fundamentally is. Downweighted differences thus cease to pose a problem. Once a model is presented as underlying reality, limitation is shifted onto reality. If materials do not fit, the question is no longer what the explanation has downweighted. Instead, reality is said not yet to have revealed the truth, the people concerned not yet to have recognized their circumstances, or relations still to remain at the surface. This allows the theory to wait forever for another layer of proof without having to answer for counterexamples that have already appeared.
Model C's ontologization is more than a theoretical declaration; it substantively changes how questions are asked. What should have been an inquiry into which differences produced this result through which relational chains is rewritten as a question of how this result displays the established structure. The former allows materials to change the explanation; the latter requires materials to supply the proof needed by the established structure. Once the boundaries are hidden, distinction extraction no longer answers for its omissions.
This self-exemption has direct practical consequences. Explanations that adopt Model C no longer need to explain why they extract only labor, property, or domination relations in a particular event, nor acknowledge that household registration, medical records, local recognition, kinship responsibilities, or national memory may change the consequences. Interpreters who adopt Model C can describe unprocessed materials as secondary, not yet manifest, or proof of a deeper total image. Explanation thereby acquires a capacity for closure.
III. Model C's Two Mechanisms
Model C contains two mechanisms that move in opposite directions but serve related functions. Total-image subsumption closes explanation through excessive certainty: a limited image produced through distinction extraction is elevated into society itself, while other valid explanations are demoted or canceled. The erasure of distinctions between explanations cancels judgment through excessive uncertainty: it illicitly substitutes the claim that society cannot be effectively explained for the fact that society cannot be exhausted and explanation is necessarily limited, then places valid models, limited models, erroneous models, and junk models on the same level. Only when complexity-based conflation uses complexity as grounds for canceling valid comparison does it constitute this mechanism for erasing distinctions between explanations.
Total-image subsumption prevents comparison from leaving the established image; the erasure of distinctions between explanations prevents comparison from beginning at all. The former co-opts counterexamples; the latter treats them as proof that no judgment can hold. The two mechanisms cannot be reduced to the same psychology, the same historical process, or the same method of argument. Both destroy the conditions for comparison, revision, and calibration.
Ordinary acknowledgment that society is complex and explanation limited does not constitute Model C. What is limited can still be effective, and compression is not necessarily dangerous. The question is whether distinction extraction preserves channels for comparison and revision. Healthy explanation must reject both total-image subsumption and the erasure of distinctions between explanations. Otherwise, it is either locked within a single image or abandons judgment amid infinite complexity.
Opposing Model C therefore does not require establishing a symmetrical 'counter-total image.' If critics say only that every generalization is dangerous, they will cast limited distinction extraction itself as suspect and end up alongside complexity-based conflation. This chapter demands a more difficult position: putting forward judgments that can be compared while openly stating which materials could revise them.
These two forms of overstepping also often borrow each other's rhetoric. Defenders of a total image can use 'reality is extremely complex' to explain any unprocessed materials and thereby postpone calibration. Complexity-based conflation, in turn, can misrepresent the demand to oppose a total image as abandonment of all judgment. Distinguishing the two prevents criticism of the total image from sliding into the thesis that judgment is impossible.
IV. The Mechanism Chain of Model C: Total-Image Subsumption
The structure of Model C in its total-image subsumption form can be unfolded through six interconnected operations: extraction of a limited set of differentiating units, fixation of weights, total-image formation, demotion of other differences, counterexample digestion, and mobilization for action. This is not a history of class theory gradually transforming from an ordinary model into Model C. It is the unfolding of class theory's basic structure as Model C in concrete arguments, institutions, and actions.
The starting point is distinction extraction within a limited scope. Abstractly speaking, extracting a set of differences from potentially infinite differences can itself be effective. Labor, wages, property, debt, housing, risk, and relations of domination can indeed enter into real consequences. When wages are overdue, labor relations, contracting chains, project timelines, and debt deadlines may immediately gain weight. This establishes only that the operation of distinction extraction must be tested against the event. It does not follow that class theory here is Model A or Model B.
Next comes the fixation of weights. Class theory presupposes that differences carrying great weight in certain relational fields universally carry the highest weight. It no longer asks whether overdue wages, a child's school admission, refusal of rental housing, or assistance for an elder are directly determined by the same set of differentiating units. Nor does it allow household registration, student registration, medical-record preservation, local reputation, or kinship responsibilities to gain weight anew in different events. Weight is thus written as a prior hierarchy instead of an intensity generated within a concrete relational field. Fixing weights also occludes the conditions under which weights change: changes in project timelines, contracting chains, residential circumstances, document validity periods, mutual recognition of medical records, and caregiving responsibilities may rapidly move previously peripheral differences to the center. Interpreters who adopt a total image need not deny these changes. They can prevent the changes from altering their original judgments simply by naming them in advance as derivative forms of the same root cause. This is the fundamental difference between fixed weights and weights within relational fields.
Then come total-image formation and demotion. Class theory and its interpreters compress the extracted differences into a class image and use that image to prescribe the ultimate form of social relations. On this basis, other differences are demoted: nation can be described as class occlusion, family as an apparatus of reproduction, local recognition as an expression of structure, and religion and culture as ideology. Such accounts need not deny the existence of each difference. Demotion is already accomplished simply by denying their independent explanatory power in advance.
Counterexample digestion and mobilization for action follow. When people with the same income encounter different consequences because of household registration, medical history, local recognition, or kinship responsibilities, interpreters who adopt class theory do not treat this as pressure for calibration. They rename the counterexamples as a deeper unfolding of the total image. Finally, organizations, institutions, and actors identify friends and enemies, allocate eligibility, deal with communities, rewrite history, or remodel families according to the total image. The new materials produced by action are then preserved and retrieved as proof that the total image was correct all along.
These six steps support one another. Fixed weights make it easier for the total image to expand; total-image formation makes demotion appear natural; demotion, in turn, leaves counterexamples without a place to enter calibration; and action write-back supplies materials to the closed structure. The connections in this six-step chain also show that Model C does not merely 'explain too much.' It has an order for processing materials: first fixing certain differences as the universal center, then allowing the remaining materials to appear only in subordinate forms, and finally turning whatever does not fit into proof. This order does not disappear because individual interpreters occasionally retain details. What matters is whether those details can genuinely change the standing of the total image. Differentiating units within a limited scope can themselves enter into and explain real consequences. Only when they are used to cancel the renewed extraction and comparison of other differences does the danger of Model C become apparent.
V. Model C and Illusory Explanatory Construction
Illusory explanatory construction is not devoid of real materials. It organizes limited materials into an image of the whole that refuses revision. Model C is its strong form, because organizations, institutions, and actors who adopt it act according to this account and leave records that can be extracted again in the future, thereby forming write-back. Action write-back, in particular, must not be described as something Model C accomplishes on its own. Concrete acts of naming, recording, mobilization, processing eligibility, and making arrangements for communities are always performed by institutions, organizations, and actors. Model C designates the closed structure through which these actions acquire direction and legitimacy in relation to one another. Restoring concrete subjects makes it possible to investigate how the structure enters real procedures, while avoiding the reduction of structural criticism to a question of personal moral character.
When organizations, institutions, and actors assign friend or enemy status, eligibility, and responsibility according to the total image, newly created records are very likely to be retrieved again as materials for the theory's self-verification. Evidential closure does not mean an absence of debate. A closed structure can absorb large quantities of material, generate intricate debates, and even allow limited adjustments to details. What it refuses is to let materials change the total image's standing. Debate then takes place only within the image: which phenomena prove it more convincingly, and which strategies are better suited to realizing it, rather than whether the image has downweighted differentiating units that are changing consequences.
The family of workers from elsewhere need not be introduced in full every time to show how this closure works. When wages are overdue, labor relations and contracting chains may indeed carry great weight. When rent is coming due, deposits, income, landlords' risk judgments, and residential stability may all enter into consequences at once. Within Model C's closed structure, these differences cannot enter calibration. Instead, they are required to return to the same total class image, as surface forms of that image in different domains rather than materials whose weights need reassessment.
Likewise, when a child enters school, household registration, residence permits, years of social-insurance coverage, student registration, available school places, and local policies may bear more directly on the event. When an elder needs assistance, medical-record preservation, mutual recognition between the place of household registration and the place of residence, kinship responsibilities, and institutional eligibility gain weight in turn. These materials are all required to return to the total class image. Local reputation, accent, neighborhood relations, historical indices of kinship, national memory, and community responsibilities may also be treated in the same way.
Counterexamples thus lose their proper function. They should force interpreters back to the relational field to ask anew what nodes have preserved, which relational chain has transmitted consequences, and which differences were previously downweighted. Yet closed explanatory procedures that adopt Model C digest counterexamples into self-verification: the harder a piece of material is to incorporate directly, the more it is said to show the total image operating at depth. Reality does not enter the explanation; reality is forced back into the same explanation.
VI. How the Erasure of Distinctions Between Explanations Cancels Comparison
The erasure of distinctions between explanations is not another error outside Model C. It is the mechanism at the other extreme through which Model C cancels comparison. Comparison does not require every explanation to use the same materials. It requires interpreters to be able to state which consequences they answer for, which materials have temporarily been placed outside the relational-field window, and when materials outside that window would constitute counterexamples. An account that fails to delimit the question yet demands that every account explain the whole has not raised the standard. It has merely made the standard impossible to meet.
Complexity-based conflation often begins with correct reminders: society is indeed complex, no explanation can exhaust reality, and any limited relational-field window may leave out differences. The problem arises when it turns these reminders into an unlimited burden of proof. As long as an explanation does not simultaneously account for nation, state, family, medical history, local relations, wages, debt, market, and historical indices, it is judged one-sided and therefore invalid.
This demand appears more rigorous, but it actually cancels comparison. A limited explanation can answer only for the key relational chains of a particular event; it cannot prove that it covers all of society. If the ability to explain everything is made a condition of validity, every explanation will fail, leaving only the conclusion that 'they are all much the same.' Complexity-based conflation does not deepen calibration. It removes calibration's starting point altogether.
Genuine calibration asks: Which differentiating units has this explanation downweighted? Has it omitted nodes? Has it cut off relational chains? Has it misread historical indices? These questions are not intended to stop judgment, but to make it revisable within a concrete relational field. An unlimited burden of proof, by contrast, rejects any provisional judgment because it demands that every judgment first complete an impossible account of the whole.
VII. Healthy Complexity Does Not Belong to Model C
Healthy humility requires interpreters to state the conditions for revision clearly: which materials suffice to support a provisional judgment, and which counterexamples, once they appear, must lead to a rewritten conclusion. It acknowledges explanation's limits without using them as a reason to evade judgment. It is willing to specify which event it is addressing, which differentiating units are gaining weight, which materials have not yet been traced clearly, and which counterexamples may change the conclusion. It retains judgment, makes explanatory boundaries public, accepts counterexamples, and takes responsibility for revision. It also preserves the possibility of withdrawal, revision, and renewed comparison.
Nor does healthy humility turn revisability into arbitrary vacillation. It does not overturn all judgments whenever a new piece of material appears. It examines what the material has actually changed: weights, nodes, relational chains, or the relational field itself. This gives revision a direction and allows limited judgment to continue bearing consequences without closing itself off. Calibration sustains precisely this kind of judgment, for which responsibility can be taken and which can be revised, rather than canceling comparison in the name of complexity.
Complexity-based conflation turns humility into a refusal to compare. It need not examine whether household registration, residence permits, years of social-insurance coverage, student registration, and available school places directly change consequences in a school-admission event. Pointing out that wages, debt, family relations, and local reputation also exist is enough for it to say that any judgment within a limited scope is merely one-sided. In this way, key differences have not been added to the analysis, yet judgment has already been forced to stop.
Respecting complexity must not flatten all differences either. It requires interpreters to acknowledge, in every event, that materials remain unextracted and to explain under which conditions those materials would change the judgment. Such revisability differs from abandoning judgment. It is precisely how limited explanation remains responsible.
VIII. The Two Mechanisms Move in Opposite Directions but Serve Related Functions
Model C's two mechanisms also create escape routes for each other. If critics counter the compression of a total image only by saying 'reality is complex,' its supporters can say that the critics have offered no judgment open to comparison. If one counters complexity-based conflation only by proposing a larger image, one falls back into subsumption. Calibration must retain both concrete judgments and boundaries that can change, so as not to shuttle between the two extremes.
Model C's mechanism of total-image subsumption and its mechanism for erasing distinctions between explanations move in opposite directions, yet undermine calibration at the same point. The former does not allow other differences independent explanatory power and swallows counterexamples into the total image, subsuming calibration. The latter does not allow valid distinctions to be established between explanations and turns counterexamples into evidence that all explanations fail, canceling calibration.
Consider again the existing materials concerning the family of workers from elsewhere. When a child enters school, household registration, residence permits, years of social-insurance coverage, student registration, available school places, and local policies may bear more directly on the event than other differences. When wages are overdue, labor relations, contracting chains, project timelines, debt deadlines, and risk-bearing may enter into consequences first.
When rental housing is refused, deposit rules, income, local reputation, accent, landlords' judgments, and neighborhood relations may work together. When an elder needs assistance, medical-record preservation, conditions for mutual recognition, kinship responsibilities, and institutional eligibility may be more crucial.
Through total-image subsumption, interpreters who adopt class theory pull these different relational fields back into the total class image in advance. The erasure of distinctions between explanations, in turn, invokes the sheer quantity of materials to refuse to explain which set of differences bears more directly on the current event. Neither mechanism has truly returned to the event. Calibration requires that return: it neither forces every question into one image in advance nor refuses comparison because materials are numerous.
IX. Calibration Must Return to Concrete Relational Fields and Relations
Calibration must return to concrete relational fields, differentiating units, weights, nodes, relational chains, historical indices, write-back, and counterexamples. These are not a fixed checklist, but entry points through which explanation can once again be examined against materials, procedures, relational positions, and real consequences. Calibration neither forces every question into a total image in advance nor refuses comparison because materials are numerous.
Calibration begins by delimiting the question and the relational field. Unpaid wages, school admission, renting, assistance, romantic relationships and marriage, migration, boundaries of national communities, and historical indices are not the same question. They involve different nodes, relational chains, preserved materials, and consequences. Delimitation does not shrink reality; it makes an explanation aware of which part of reality it is tracing.
The relational-field window must also be constrained by materials. If it is too narrow, differences already operating along relational chains will be mistaken for background. If it expands without limit, it falls into the mechanism that erases distinctions between explanations by canceling valid comparison in the name of complexity, leaving no judgment a place to stand. Delimiting the question is not arbitrary cropping. It establishes provisional boundaries according to the nodes, materials, and consequences already implicated in the event, with readiness to adjust them when counterexamples appear.
Delimiting the relational field also means distinguishing the roles of the same material in different events. Household registration may directly restrict eligibility in school admission, act indirectly through conditions for mutual recognition in assistance, and not necessarily lie on the main relational chain in an unpaid-wages event. Without this distinction, interpreters can readily misrepresent a piece of material's real weight as its fixed domination over all events.
In an unpaid-wages event, contracts, wage slips, construction-site records, contracting chains, project timelines, and debt deadlines may form the principal entry points. In school admission, household registration, residence permits, years of social-insurance coverage, student registration, available school places, and local policies may bear more directly on the event. Renting and assistance, in turn, bring deposit rules, medical records, conditions for mutual recognition, kinship responsibilities, local reputation, and institutional eligibility into different relational chains. These differences cannot be arbitrarily spliced into a list. Whether they gain weight depends on actual consequences within the event.
Without a delimited question, the mechanism of total-image subsumption can pull all events back into the established image. The mechanism that erases distinctions between explanations by canceling valid comparison in the name of complexity, in turn, declares all events incomparable. The relational field deprives both forms of overstepping of their excuse: it requires interpreters to specify a concrete state of relations instead of replacing the event with a grand name.
Comparison must also distinguish 'being downweighted' from 'not yet having gained weight.' The latter does not necessarily mean omission; a difference may simply not yet have been triggered by an event. The former means that, within an explanation, a difference's weight, visibility, opportunity to enter comparison, or access to explanatory standing is compressed, restricted, or excluded. These states must not be conflated. Whether a difference has actually entered into consequences, whether an explanation acknowledges or traces it, whether its current weight is relatively low, and whether it has been excluded from explanation all require separate judgments. The problem with Model C is precisely that it often directly demotes differences not absorbed by the total image to the surface level. Healthy explanation preserves the possibility that they may gain weight anew, while not misrepresenting that possibility as a current conclusion.
Nor must a downweighted difference disappear entirely before it deserves examination. It may not yet have produced visible consequences, even though it has already been preserved at nodes and is accumulating along relational chains. Calibration must recognize this possibility without elevating all materials that have not yet gained weight into the current center. This is precisely why the relational-field window must be limited yet capable of adjustment under pressure from counterexamples.
Calibration must also compare what different explanations have extracted and downweighted in the current event. Labor, debt, housing, and risk can form one set of entry points; household registration, documents, archives, and institutional implementation form another. Local relations, accent, neighborhood recognition, and community boundaries, as well as kinship responsibilities, medical history, historical indices, and care arrangements, may also change consequences. Arranging these names into lists of equal length does not complete comparison. Explanation must still follow the concrete event and trace how each difference enters into consequences.
After a family experiences unpaid wages, labor relations and debt deadlines may gain weight first. Once rent payments break down, rental nodes change in turn. If residential instability affects a child's school admission, household registration, residence permits, student registration, and school places acquire new positions. When medical records cannot be mutually recognized, the preservation of medical history and institutional eligibility may become constraints in turn. Differentiating units change weight along nodes and relational chains; they do not obey a total image fixed in advance.
Comparing explanations therefore does not mean finding class theory a place in an ordinary toolbox. Some differentiating units extracted by class theory may have explanatory value within a limited scope; class theory itself remains Model C. Calibration compares whether concrete differences, nodes, and relational chains actually change consequences. It does not give class, state, market, nation, or family a share of society apiece.
Weight is not a rank that a difference carries by nature. It is the intensity with which the difference enters into consequences within a concrete relational field. When wages are overdue, low income may occupy the center. Contracts, wage slips, construction-site records, and coworkers' testimony can preserve and certify differences involving labor. Overdue wages may reduce the capacity to pay rent.
When a child enters school, household registration, residence permits, years of social-insurance coverage, and student-registration nodes may bear more directly on the event. Administrative records, school systems, and local policies process household registration and student registration. Residential instability may make school-admission materials difficult to maintain, and interrupted social-insurance contributions may change eligibility judgments.
When an elder needs assistance, medical-record preservation, institutional eligibility, and kinship responsibilities may change the result in turn. Hospital medical records, medical-insurance systems, hospitals in the place of origin, and assistance procedures in the place of residence process medical history. How medical records are preserved may affect whether assistance can be certified.
Historical indices of kinship may affect borrowing and caregiving responsibilities. Neighborhood relations, landlords' impressions, networks of people from the same place of origin, and community reputation preserve local recognition.
A node is not an abstract dot. One must explain which differences it has processed and how it has brought them into consequences. Relational chains change the forms and weights of differences in transmission and seldom exhibit smooth linear causation.
Reassessing weights, examining nodes, and tracing relational chains cannot substitute for one another. If one says only that a difference is weighty without knowing where it is located or who certifies it, the judgment cannot be grounded in real procedures. If one merely lists nodes without tracing transmission, materials remain on a list. If one traces only a single chain without reassessing weights, partial causation is readily misrepresented as the entire cause. Together, the three constrain explanation's impulse to expand arbitrarily.
These three checks can therefore distinguish explanatory power from the imposing sound of a name. A grand name may encompass many materials without explaining at which node those materials are formatted or how they change eligibility along relational chains. Conversely, an apparently small record, condition for mutual recognition, or instance of local recognition must not be downweighted for lacking a grand name if it sits precisely at a gateway to consequences. Using only a total image swallows these changes; the mechanism that erases distinctions between explanations by canceling valid comparison in the name of complexity abandons the effort to trace them.
Calibration must also examine historical indices. Past differences do not automatically govern the present. They must be located and extracted anew through current events, memory, archives, institutional records, local reputation, or community boundaries. A piece of kinship history may ordinarily be inconspicuous, yet it gains weight when a family borrows money. Memories of migration may become materials for judging stability and trustworthiness in school admission, renting, or neighborhood conflicts.
Examining historical indices can prevent two opposing simplifications: presenting the past as forever determining the present, and treating any past as irrelevant background. Neither investigates mechanisms of preservation. Family memory, medical records, local reputation, religious recognition, or national commemorations change people's eligibility, responsibilities, or risks in the present only when they are extracted anew, certified, and transmitted at concrete nodes.
Old local events, national memory, religious recognition, and community responsibilities do not bear on all events with the same intensity either. Only when nodes preserve them anew, certify them, or transmit them within particular relational fields can they change present consequences. A historical index is not a total image that hands the present over to the past. It requires asking in what form, through whom, and along which relational chain the past returns to this moment.
Finally, counterexamples must be allowed to enter explanation. If a counterexample shows that an important set of differentiating units has been downweighted, interpreters must extract those differences anew. If it shows that a judgment of weight was miscalibrated, they must reassess weights. If it shows that the relational field was wrongly delimited, they must return to the event. If it shows that nodes or relational chains have not been traced clearly, they must follow up with the tracing still needed.
Counterexamples also matter because they force interpreters to acknowledge that consequences may arise from intersections not previously seen. A counterexample does more than add a new factor. It may show that the same differentiating unit is reformatted through different nodes, or that a relational chain previously cut off is in fact still operating. Such reopening can change an explanation more than adding a footnote beside an existing image. Total-image subsumption prevents counterexamples from entering calibration. The mechanism that erases distinctions between explanations by canceling valid comparison in the name of complexity uses counterexamples to declare every judgment a failure. Calibration rejects both treatments.
X. Calibration Is Not Compromise, nor Does It Reserve a Place for Class Theory as a Tool
Calibration's rigor lies precisely in allowing uneven conclusions. In one event, labor relations may bear more directly on consequences than local reputation. In another, local recognition, student registration, or medical-record preservation may change a judgment originally inferred from income and debt. There is no need to favor any one domain, nor is an even distribution sought. Consequences, nodes, and relational chains determine the direction of comparison. As long as a total image can still decide the answer in advance, such an uneven judgment cannot genuinely occur.
Explanatory calibration rejects moderate compromise. Nor does it list class, state, market, nation, family, culture, and history in turn and then say they all have some truth to them. Compromise places different explanations on an equal footing without answering which differentiating unit is changing consequences at this moment, which node has preserved key materials, which relational chain has been cut off, or which counterexample would suffice to change the judgment.
Calibration can establish that certain differentiating units extracted by class theory carry high weights in the current event. This judgment concerns concrete differences; it does not retain the total class image. In unpaid-wages cases, labor relations, contracting chains, project timelines, and debt deadlines may need to be traced first. This does not mean that the class image acquires the standing of an ordinary limited tool, still less that class theory can be placed alongside Model A and Model B and freely selected according to the event. Class theory remains Model C. Likewise, household registration, medical records, local recognition, kinship responsibilities, or historical indices gaining weight in another event gives no reason to allot a share to every grand concept. Calibration does not retain the class framework and then add other variables. It requires a return to the concrete relational field each time, to judge which differentiating units, nodes, and relational chains actually change consequences and to let counterexamples revise that judgment.
It is precisely these two illicit substitutions that calibration rejects. One account appears to acknowledge that labor, wages, housing, and debt carry great weight in certain events, then immediately demands that the class image be retained as the overall framework, with household registration, locality, family, or nation added to it. This is still not calibration. It decides the total image for the event in advance, allowing only some differences to be added or removed at its edges. Calibration does not accept an arrangement in which the center is fixed first and materials supplied afterward.
Another account lays all materials out flat: labor relations, household registration, medical history, local recognition, kinship responsibilities, and national memory are all present, so each is said to be equally important. This too evades judgment. Differences can coexist without entering into consequences at the same intensity within the same relational field. Calibration neither gives them an eternal ranking nor refuses to rank them. It asks which are changing eligibility, risk, responsibility, and action space.
Calibration can neither remove class theory from its position as Model C nor establish any other grand concept as the permanent center in advance. It addresses how concrete differences are preserved by nodes, transmitted through relational chains, and certified by institutions, and how they change existing judgments when counterexamples appear. This work does not provide a common answer for every domain, but requires every answer to state its materials and boundaries.
XI. Bringing the Theory Together and the Responsibility of Application
This restriction also allows theoretical criticism to reach concrete consequences: instead of asking who holds the sole key, it asks what kinds of naming, recording, and handling are pushing people into which differentiation positions and what materials they leave that can be extracted again in the future.
Such self-limitation does not reduce Social Differentiation Network Theory to a neutral list of tools. It still advances the judgment that social consequences require tracing differentiating units, weights, nodes, relational chains, historical indices, and write-back within concrete relational fields. But it does not allow this judgment to subsume any event in advance. A theory's strength should come from its capacity to be moved by materials, not from its refusal to let materials change it.
Limited explanation cannot do without simplification. If simplification occludes key differences and forgets its own distinction extraction and what it has downweighted, it forms dangerous compression. Once dangerous compression stabilizes, distinction extraction, weights, relational fields, nodes, relational chains, and historical indices develop differentiation bias in relation to real consequences. Once that bias closes itself off through gap completion, causal-line infilling, counterexample digestion, and action write-back, it forms an illusory explanatory construction. Model C's total-image subsumption and erasure of distinctions between explanations are two opposing forms of overstepping along this chain. The former conflates differentiating units effective within a limited scope with the total image, subsuming other valid explanations. The latter illicitly substitutes equal failure for limitation, canceling comparison, judgment, and calibration. They should neither be reconciled through compromise nor reduced to each other. Maintaining the responsibility of limited explanation lies precisely in rejecting both extremes.
Social Differentiation Network Theory, too, must not on this basis claim to be a new underlying reality or total image. It must not reject counterexamples or demote concrete relational fields, nodes, and relational chains to illustrations of what it already knows. In every analysis, it must face the same questions: What has been extracted? What has been downweighted? Which consequences have not yet been traced clearly? Which materials would force it to change? If it loses these constraints, it too will degenerate into a new Model C.
What application retains is this responsibility, not a scheme and a set of answers that precede the question, nor a reproducible template or formula. Application is therefore also a test of the theory's limits. If, when faced with concrete materials, a theory can only repeat existing names and cannot explain why the materials gain weight at this moment, how nodes process them, or which counterexamples might bring about change, it has not truly entered reality. Conversely, a concrete question need not provide proof that a particular theory is forever correct. It requires only that interpreters answer for the judgments they make.
Different domains do not reveal problems in the same order. Some problems first appear between caregiving responsibilities and medical-record preservation; others, between household-registration eligibility and available school places; still others, between labor records, contracting chains, and debt deadlines. The responsibility of application is to begin from their respective contradictions and consequences, rather than turn any domain into an exercise in the same set of concepts.
Application requires taking on calibration anew in different questions: how differentiating units gain weight, how weights are misjudged, which nodes preserve key differences, which relational chains are transmitting consequences, which historical indices are triggered anew, and which counterexamples require explanation to change. Pasting a theory directly onto reality, or using concrete domains to provide self-verification for this book, departs from this responsibility.
This way of entering reality neither allows a total image to answer for the event in advance nor stops judgment on grounds of complexity. Only when a theory can still explain its boundaries, accept counterexamples, and revise its distinction extraction in concrete questions does it remain within its own limits.