Chapter 6. Concepts Are Not What Society Fundamentally Is: How Concepts Compress Differentiating Units and Occlude Reality
I. Concepts Are Not the Enemy
Without concepts, experience is difficult to remember, compare, and discuss. Wages, employment, debt, rent, medical records, household registration, accents, cash gifts, kinship relations, real estate, educational qualifications, religion, the body, age, prior relations, school places, evidence of workplace injury, and family responsibilities would remain scattered across individual experiences if there were no names at all by which to organize them. People in conversation first draw on names such as “wages,” “the state,” “the market,” “culture,” or “the family” to identify the general direction of the material: labor and income; law and registration; prices and credit; rites and customs, and memory; care and dependence. Only by using names to organize this material do people sketch a provisional outline.
Through names, people remember recurring relations and compare experiences separated from one another. Without this step, public judgment can readily fall back into scattered impressions and individual stories. People also use concepts to transmit material for discussion. When someone describes a difficulty with school admission, a case of unpaid wages, or a cash gift, they need not retell every detail each time. Listeners can first use the name to grasp the general direction of the problem, then return to the material to see whether that name applies. Concepts, therefore, are not the enemy. The problem begins when names are used in place of further inquiry. If a name serves only as an entry point, it can still lead people back to labor, documents, time, care, records, and the ways things are taken up and sustained; if it is treated as an endpoint, the material begins to disappear behind the name.
This distinction cannot be applied directly to class theory. Labor positions, wages, property, employment relations, debt, housing, resource buffers, the bearing of risk, the ability to exit, and domination are all real differences that carry considerable weight, and each can enter a bounded analysis. Class theory, however, organizes these differences into a total class image from the outset. It is not an ordinary concept that was originally safe and only later lost its boundaries.
General concepts can leave room to be unfolded again. The structure of class theory, by contrast, compresses, projects, and names the extracted differences as a total image, making it easy for other differences to be arranged inside that image. The fact that it grasps some weighty relations does not justify preserving it as a safe entry point.
Conceptual compression has at least three distinct levels. The first is naming for communication: gathering scattered events into formulations that can be remembered. The second places material within a particular explanatory framework and indicates which differences are considered important. The third may treat the framework, in reverse, as the pre-existing structure of reality and require all subsequent material to prove that it is merely an expression of that structure. The first two levels can be useful; it is the third that constitutes the ontologization of concepts.
II. How Concepts Compress Differentiating Units
A concept is a name into which people compress a set of related differentiating units in order to discuss, remember, compare, and judge reality.
Compression here is not an action that a word performs on its own. In naming, classifying, registering, and discussing, people, interpreters, organizations, and institutional procedures extract some differences whose relations have already become apparent and group them under a name. The name is both the result of this compression and an entry point for subsequent use. What is compressed is not empty abstraction, but concrete differences and the relations they have already manifested.
This selection has a different focus in different concepts. In speaking of “the state,” users may bring together law, registration, household registration, school places, public finance, administrative boundaries, public resources, implementation at service counters, and supporting material. In speaking of “the market,” they may bring together prices, wages, rent, credit, exchange, competition, the ability to pay, risk, and opportunity costs. Likewise, users draw on “culture” to organize rites and customs, forms of address, face, gender expectations, memory, local habits, taboos, and judgments of closeness and distance. When naming something as “family,” they provisionally group together care, dependence, inheritance, debt, favoritism, prior relations, eldercare, intimacy, control, and available exits. The material drawn on need not be exactly the same in every case, yet people readily regard it as naturally belonging to the same kind.
Compression preserves similarities among certain differences, as well as the experience of their having jointly acquired greater weight in some problems, so that people do not have to name every encounter from scratch. Material included under a name becomes easier to see, while material that has not been extracted may recede into the background. Reality has not disappeared; explanation, argument, and institutional handling may simply no longer give it the same place.
A name must not be conflated with the material compressed into it. “The state” is not registration, public finance, and service counters themselves, nor is “the family” identical with care, debt, and prior relations. People use names to gather material provisionally, but in another matter that material may still connect to different relations and consequences. Nor can the same name guarantee that the differences within it have the same operative weight: household registration and school places form thresholds in different ways, just as housing prices and wages alter choices in different ways. People use names to sketch an outline; detailed judgments must still return to the material. The possibility of unfolding the material again is precisely a condition of a concept's limited validity. Users should explain what they have extracted at this point and acknowledge what they have omitted. When counterexamples, boundaries, and new material enter, the original compression can be corrected, rather than reality being required to accommodate the old name.
Concepts can be used repeatedly because an act of compression leaves a name that can be taken up by others. Later users need not have experienced the original relations themselves to approach, through this name, experience that has already been organized. Yet carrying a name forward is not the same as carrying forward all the material. The more stable the name, the easier it becomes to forget how the original material was selected. Compression also changes the scale of comparison. Once people place several differences under one name, they often find it easier to see what those differences have in common, but harder to see their respective times and origins, and the conditions under which they are taken up and sustained. Comparison thus gains an entry point, and also a point at which distortion may occur. It is precisely these differences, hidden by the common outline, that need to be unfolded again.
Naming practices are not always carried out by the same kinds of actors. Everyday conversation, research explanations, organizational classification, and institutional registration gather material in different ways; what is retained and what is omitted may also differ when the same name enters different procedures. Naming itself does not change life unless someone connects the name to records, qualifications, authority, resources, or responsibilities. The term “workplace injury” may have different consequences in procedures involving medical records, assessments, compensation, and leave; the term “nation” may likewise carry different relations in family memory, education, legal registration, and the handling of boundaries. A concept's efficacy comes from how specific people and procedures use it, rather than from a fixed power possessed by the name itself. Differences in authority also arise when different participants use a concept. Those involved can use a name to describe their experience, staff may use fields on a form to connect the name to qualifications, and researchers use it to organize comparisons. The three have different capacities to produce consequences. The fact that the word is the same does not justify assuming that all users are performing the same action.
Concepts also change shape as they are passed between stages. A school understands “residence” as material bearing on eligibility, a community as long-term interaction, and a family as a fact of everyday life. The same word enters different consequences at different nodes. If we compare only the wording, we can easily mistake different forms of handling for the same judgment. What must be traced, therefore, is which node extracts the concept, what format it adopts, and to whom it transmits the result. Problems of translation across contexts are connected to this as well: legal, family, local, and academic language may use different names to describe the same part of an experience, or the same name to handle different relations. Whether translation succeeds depends on whether material, responsibility, and consequences can be connected. Merely aligning words does not prove that a concept has the same meaning at different nodes.
The historicity of concepts must also be preserved. The same name may be supported by different material in different institutions, generations, and local relations. Continuity of the name does not mean that the mechanisms of consequences have remained unchanged. Names for nations can carry forward long-term memory, but may also be reformatted by institutions; the name “family” can sustain kinship responsibilities, but may change meaning through household division, migration, and conflicts over care. Concepts must be traced together with the relations that formed them.
Boundary cases often expose the conflation of concepts: an organization may both participate in market transactions and receive authorization from the state; a family may both preserve community memory and handle property arrangements. The same object can thus enter different relations. Analysis must explain the material, authority, and consequences of each relation separately. The fact that the object is the same does not justify combining the concepts of market, state, family, or community into a single actor.
III. Why Concepts Have Power
Through shared names, people bring widely separated experiences into relations that allow comparison. Events need not remain forever a matter of private experience alone. If a person's difficulty is always confined to a private account, it is not easy to establish comparisons with other people's material. Names bring situations that were previously difficult to communicate into public language. In a case of unpaid wages, wage arrears, rent pressure, lack of savings, the cost of defending one's rights, alternative employment opportunities, and domination in labor can be seen together. In school admission, household registration, student registration, school places, certificates from subdistrict offices, education department rules, and implementation at service counters can also be placed within the same set of institutional relations. In cash-gift exchanges, records, social respectability, closeness and distance, seating arrangements, prior relations, and future mutual aid can enter discussion. This organization speeds up public judgment. People can first see which problems are not isolated accidents, then ask through which concrete stages they are connected. The practices of naming and organizing material keep discussion from being reduced to “I had bad luck” or “He was unlucky.”
People, organizations, and institutional procedures also use names to preserve experience. Repeatedly used names retain in memory a set of relations that once appeared together. Later users draw on these names to approach old problems, and organizations and institutional procedures often use classifications to record, compare, and handle material. Institutional handling depends on names, fields, and classifications to preserve records, compare qualifications, and make decisions. Precisely for this reason, the use of names is never merely a question of language: it affects which material can readily enter a procedure, which requires additional explanation, and which experiences have no existing field in which to be placed at all. Concepts may also harden through repeated institutional use. When the same field, the same classification, and the same set of evidentiary requirements are used across multiple events, people prepare the corresponding material in advance, and experiences that do not fit the format find it harder to enter judgment. This stability is a real consequence, but it still comes from the ongoing handling by rule-makers, record-keepers, reviewers, and those implementing decisions. It does not mean that the concept itself has become an entity capable of acting.
Yet preservation is never complete preservation. The easier a name is to transmit, the more easily the disagreements, conditions, and exceptions once within it are occluded. Compression brings convenience, but also limits. Public judgment depends especially on concepts, yet cannot stop at them. After saying that a rule is an administrative matter, that an expenditure involves market pressure, or that a set of interactions involves local rites and customs, one must still return to specific documents, deadlines, the ability to pay, and prior relations to see how they alter consequences. The power of a name comes not from its standing above reality, but from users being able to enter reality through it. Apart from material, procedures, and relations, a name is nothing more than a slogan to be fought over.
IV. How Concepts Occlude Reality
Occlusion does not mean that reality disappears. Reality continues to unfold, but often it can enter explanation, records, and public discussion only through classifications already in use. Those who use class theory may see more of labor, wages, property, debt, housing, and domination, while downweighting language, religion, the body, local memory, prior relations within families, rites and customs, and institutional identity. Those who use the name “the state” may likewise focus on law, registration, and administration, while placing rent, help from relatives, market costs, local familiarity, and family care in the background. The first set of material is not false, nor is the second set real only on that account. If interpreters forget their own choices about what to retain and omit, they will mistake the differences already extracted for the whole of reality.
Within a market framework, cash gifts, support for elderly relatives, family affection, and local credit are often treated as costs and returns. When interpreters adopt a cultural framework, housing prices, childcare costs, gendered risks, and the pressures of eldercare are readily assigned to traditional beliefs; when they turn to a family framework, unemployment, debt, housing costs, and industry cycles may in turn be classified as private trauma. Each account may grasp some relations and omit others. Occlusion also comes from institutional classification. A procedure specifies which forms of proof count, which experiences can be registered, and which relations can remain only in oral accounts. It is not concepts blocking material on their own: organizations, institutional procedures, and naming practices place some material in visible fields and leave other material outside them. Once existing classifications become the only language, unclassified differences find it difficult to gain standing in an explanation. They may still affect qualifications, waiting, responsibilities, and action, yet find it difficult to enter formal judgment.
Occlusion is often not deliberate concealment. Interpreters may sincerely organize material under names familiar to them, while institutional procedures may simply continue using existing fields. As names are used repeatedly, differences not included come to look more and more like irrelevant facts, even though they continue to alter people's choices and consequences. The same material also takes on different appearances under different names: debt can be seen as market pressure, a family burden, or a problem of institutional eligibility. Any of these accounts may grasp some of the facts. If only one is permitted to exist, the other relations lose the opportunity to be examined.
Names for nations must also be treated separately here. Names can be compressed, invoked, classified, or put to use; real national communities, however, cannot on that account be reduced to mere concepts, identity labels, state registration, subjective imaginings, or appendages of class. The compressibility of names does not cancel the real relations through which communities have formed historically.
Placing class, the state, market, nation, family, and culture side by side does not automatically yield a complete explanation. Institutions may also create records according to classifications and then use those records, in reverse, to prove that the classifications are correct. Both practices readily allow names to occlude the actual handling. Analysis must trace who classified first, which fields and rules preserved the results, how subsequent personnel retrieved them, and whether contrary material can correct existing records. Only by opening up this formation chain can the juxtaposition of multiple concepts and institutional records be kept from replacing causation, time, and responsibility.
V. How Concepts Come to Be Treated as What Society Fundamentally Is
When a concept serves as an entry point, users still return to the material it indicates. If judgment stops at the name and says only that this is a state problem, a market problem, a cultural problem, or a family problem, documents, time, relations, the ways things are taken up and sustained, and responsibility are no longer investigated. Once a name is treated as fundamental reality, explanation undergoes an inversion: a judgment that should have been supported by material instead requires the material to prove the name. Relations that do not fit the name are described as secondary, exceptional, or superficial, and differences within the name are forcibly arranged in the same direction. Concepts do not inherently produce this result on their own; the problem lies in users and procedures refusing to unfold them again.
The ontologization of general concepts occurs when names are mistaken for independently existing social entities. Interpreters no longer explain who extracted which material or which procedures gave it operative weight, and no longer allow new material to change the judgment. A name changes from a tool for organizing experience into something that adjudicates experience in advance. When names seem able to arrange reality on their own, the actual actors who extract, classify, retain, omit, name, and adjudicate are erased. These actions must always be traced to specific interpreters, organizations, institutional procedures, and classificatory practices. Only by restoring the actions to these actors can we see where a judgment begins to lose its boundaries. The state, market, culture, and family can all be misused in this way. They can point to entry points into different relations, but cannot declare a sole cause on reality's behalf, still less reduce all other relations to their own appendages. General concepts may still be brought back within bounds by boundaries, counterexamples, and material. If users explain the scope of their material, acknowledge omissions, and allow judgments to contract, names will not readily replace society itself. What is at issue here is the limited use of general concepts, not an exemption for class theory.
Class theory requires separate treatment. It is not an ordinary concept that turned from an entry point into a wall only after later misuse. From the outset, it compresses, projects, names, and organizes differences such as labor positions, wages, property, employment relations, debt, housing, resource buffers, the bearing of risk, the ability to exit, and domination into a total class image. These differences can each enter a bounded analysis. The total class image itself, however, cannot be preserved as a limited entry point, with the state, market, nation, culture, and family then treated as peripheral supplements. Its capacity to extract differences within a limited scope has not changed its basic structure of subsuming explanations.
Abstract names cannot bear responsibility on behalf of actors. The market does not set prices on its own, the state does not carry out registration on its own, the family does not allocate care through a unified will, and the nation does not preserve memory on its own. Specific participants, organizational positions, rules, and procedures are the carriers through which differences are handled. Concepts can generalize their relations, but cannot become the source of their actions.
The power and danger of concepts come from the same capacity for compression. A concept organizes scattered consequences into comparable problems, but may also fix the outline obtained through comparison as a pre-existing world. Work with concepts therefore does not aim to find the largest name. It must maintain distinctions among naming, extraction, weight assignment, passing on, and write-back, so that material can reopen generalizations. Only under this condition is a concept a bounded explanation, rather than an image elevated into what society fundamentally is. Concepts also function differently in different settings: in public language a concept may gather experiences together; in fields on institutional forms it can connect to eligibility and implementation; in academic discussion, research concepts are used to organize comparisons. Even when all three use the same name, their respective carriers and consequences must be distinguished. Continuity of wording cannot stand in for continuity of relations.
Saying that concepts are not what society fundamentally is does not negate their power. Generalizations can still organize real problems, but must explain who carries them, when they form, on what material they are based, through which nodes they pass, and what consequences they produce. They must also acknowledge what contrary material would invalidate the conclusion. A concept's force comes from its ability to accept limits in reality while continuing to explain, rather than from securing a position immune to correction. When a concept proves effective repeatedly across events, an analyst can strengthen the judgment concerning its pattern of relations. Yet this stability is still supported by material and relaying; it does not elevate a generalization into an entity capable of acting on its own.
The boundaries of concepts must ultimately be grounded in responsibility. If “institutions,” “the market,” or “culture” leave people unable to identify who makes rules, allocates resources, preserves material, rejects objections, implements consequences, or holds the authority to correct, abstract names have already taken over the actions of actors. Whether analysts use these names to explain or to criticize, they should restore control, knowledge, authorization, and the capacity to revise to concrete positions. Disaggregating them in this way does not lessen criticism; it enables criticism to reach the relations that produce and sustain consequences.
VI. Differentiating Units Must Not Be Ontologized Either
Differentiating units cannot be elevated into what society fundamentally is under a new name, nor should they be treated as social atoms, ultimate entities, or an all-purpose term. They cannot replace class, the state, market, nation, culture, and family as another, still larger total concept. Breaking concepts back down into differences does not mean that we have found society's final material. Only within concrete problems, as differences connect with other relations, does it become apparent what they can change. Detached from these processes, arranging differentiating units into a fixed list would likewise freeze analysis back into a plane.
Treating differentiating units as new labels merely gives the old practice a different word. If someone says, “This is a problem of differentiating units,” but no longer explains which differences are extracted, by whom they are recognized, how they are transmitted, and when they enter consequences, the name likewise becomes empty. Nor can differentiating units serve as an all-purpose explanatory key. They remind people to ask how differences are extracted, recognized, and transmitted, but do not guarantee that every difference named will enter consequences. Differentiating units acquire meaning only within concrete problems. Which relational field they enter, which relations they connect with, and at which nodes they are preserved, certified, or passed on are what enable certain differences to enter qualifications, responsibility, action, or consequences. Nor can they be detached from relations of operative weight: a difference that is prominent in one matter may recede into the background in another, and the same record may produce different consequences when adopted by different procedures.
The judgment that “society is a differentiation network” has real standing in social ontology and cannot be reduced to an analytical method, an observational perspective, or a rhetorical metaphor. Equally, users cannot expand it into an ultimate entity or highest total image that is closed, exclusive, and completed in advance, that encompasses everything that exists in the universe, and that is immune to counterexamples. How concrete differences are extracted, connected, assigned operative weight, recognized, and adjudicated must still be traced to interpreters, organizations, institutional procedures, and actual relational processes. The network is neither a mechanical aggregate nor an actor operating on its own. Preserving this boundary serves to maintain the differentiation network's reality, the way society is formed, and the network's openness together; it does not withdraw this book's affirmative proposition in social ontology.
Criticism of concepts must also delimit its target. Pointing out that a particular usage occludes reality does not cancel the name's function in every setting. Rejecting class as what society fundamentally is still leaves room for limited analysis of wages, property, debt, and domination. Analysts should specify whether they are criticizing ambiguity in the name, the explanatory structure, institutional usage, or the actions of specific people, so that criticism does not create yet another undifferentiated abstract object.
VII. How to Use Concepts Properly
Using a concept properly requires first explaining which material it points to at this moment. In speaking of the state, does one mean legal texts, household registration, registration, school places, implementation at service counters, or administrative boundaries? In speaking of the market, does one mean prices, wages, rent, credit, competition, or opportunity costs? A name cannot replace this explanation. One must also explain what the compression has retained and what it has omitted. If a judgment sets out only the material it has seen, readers cannot know which relations it has omitted or whether those omissions would change the conclusion.
The scope of a concept's application should also be stated. A name being helpful for one kind of problem does not mean that it can cover all problems; its grasping a key relation at one moment does not mean that the same relation will always retain the same operative weight. A concept's valid scope may change with the research question, but cannot be moved arbitrarily after the fact. Analysts should state their material and scope at the outset. If adjustment is needed after counterexamples appear, they must make public which participants, nodes, or consequences have changed. Merely excluding unfavorable cases from a new definition does not count as conceptual calibration.
The omissions made by concepts are sometimes necessary, but what has been omitted must be capable of being unfolded again. Without compression, analysis cannot form communicable judgments. Yet readers should be able to ask what material lies beneath a concept, which relations have been downweighted, which conclusion holds only for a particular event, and what new counterexample would deprive the concept of explanatory power. This means that users must be able to unfold again the concrete material contained in a name: labor, wages, debt, housing, and risk can return to their respective problems, while law, registration, public finance, service counters, and public resources can each be examined separately. Unfolding again still preserves a concept's limited uses. Its purpose is to prevent concepts from adjudicating in advance on reality's behalf. A limited concept should remain correctable: it can aid memory, comparison, transmission, and public judgment, but must leave an exit back to concrete material.
A concept's clarity depends on whether the basis of its formation can be traced. If, after a macro-level judgment returns to material, nodes, and relational chains, it is no longer apparent how the name was formed through extraction, weight assignment, and consequences, compression may have gone too far. If the core inference can still be re-examined by following the material, the concept remains open to inspection. This does not require writing to reproduce all of reality. It requires that names not allow an argument to leap over stages that have yet to be explained.
In the case of class theory, one cannot say that “there is no problem with using the concept of class.” What can enter analysis are concrete differences such as labor positions, wages, property, employment, debt, housing, resource buffers, the bearing of risk, the ability to exit, and domination. Returning these differences to their respective problems will not assemble them into a safe total class image.
Reopening concepts does not mean abandoning generalization. Analysts may retain “the family,” “the market,” “the state,” and “the nation” as provisional entry points, but must indicate the material they currently cover, what remains outside them, and the time span within which they apply. When counterexamples show that an entry point is too coarse, analysts should add differentiating units, split nodes, or narrow the conclusion, rather than wrap the problem up again in new grand words. When multiple concepts intersect, one should not establish a still higher total term to gather them all in. Family, state, market, and nation can affect one another within the same event; analysts must explain through which people, material, and procedures each enters consequences. The intersection of relations supports multiple traceable paths, not a new omnipotent concept.
The use of concepts must accommodate counterexamples. When new material appears, interpreters must be able to narrow, adjust, or abandon their original judgments. Names cannot become shields against revision. When new records, relations, or consequences incompatible with the original judgment appear, users must address them directly, rather than dismiss them all as exceptions. Whether a concept has been overextended can be examined precisely through boundary cases: when an event fits the name in only some of its material while other relations point to different outcomes, can the interpreter acknowledge that the classification does not fully apply? When two events bearing the same name produce different consequences, can the concept be broken back down into nodes and relational chains? If every difference can only be gathered back under the same name, the concept loses the boundaries through which reality constrains it.
Whether a concept can return to reality ultimately depends on whether it allows material to change the concept's own organization. If new evidence is merely added to an existing framework without being able to change weights, paths, or scope, its supposed openness is merely formal. A concept that can truly be calibrated leaves room for revision: it can be narrowed, differentiated into levels, or given a different entry point, and it can be withdrawn if it continues to fail.
Concepts can form a public language that allows scattered events to be compared and downweighted experiences to enter discussion. This function does not require all material to submit to the same name. New records may break up existing generalizations, and other research can pose different questions on the basis of law, history, institutions, or the experiences of those involved. As long as the shared language still allows participants to add material, dispute the scope, and change judgments, concepts can enter comparison. If the unknown, objections, and outside research can only be taken in as examples of the concept itself, it begins to lose its limited boundaries. Concepts remain necessary theoretical tools. They allow different events to be compared and long-term relations to enter an explanation. The greater their power to generalize, the more they must be able to return to concrete relations and indicate what material they have retained, what differences they have downweighted, and under which conditions they offer only a partial explanation. Concepts are therefore neither temporary labels nor ultimate answers, but forms of argument constrained by the boundaries of reality and the conditions under which they fail.
VIII. Turning to How Differences Produce Consequences for Action
Once concepts have been reopened, concrete analysis must avoid the closure of grand concepts while also avoiding listing differences as unrelated fragments and breaking every person down into countless unmanageable details. The question is no longer how to add more names to the same matter, but how concrete differences move from the background into qualifications, positions, responsibility, action, and consequences. People can certainly use concepts to explain reality. Whether that account can withstand correction by reality, however, depends precisely on how concrete differences enter consequences: analysis must explain which differences in the problem at hand have already acquired operative weight, how they enter consequences along relations, and what material might change the judgment. Consequences always take shape at points of connection and continuation. What must be asked is which differences have already connected with institutional procedures, the taking up and sustaining of differences through relations, and temporal conditions, thereby enabling people to act, requiring them to wait, allowing them to exit, or forcing them to bear consequences.
These questions all require tracing the places where differences connect with procedures, relations, and the ways differences are taken up and sustained: under what conditions a residence permit changes eligibility for school admission; how surveillance footage enters the determination of responsibility for workplace injury; how records of wedding and funeral exchanges affect local reputation; how accents connect with closeness and exclusion; and how debt deadlines change whether people can afford to resign, marry, be ill, or refuse work. Differences do not produce consequences as soon as they appear. Some material must be registered, some recognized, some transmitted through relations, and some supported during a period of waiting before it can alter the room available for action. When material is recognized, differences may enter qualifications; when records are preserved, differences may reappear later; when relations take differences up and sustain them, or refuse to do so, differences change the room available for action. Nor do the same differences acquire equal weight in every problem. When work arrangements, family care, institutional deadlines, local relations, and prior records connect with one another, material that was previously inconspicuous may suddenly affect choices; material that appears weighty may also temporarily fail to enter the outcome because it has not been taken up and sustained.