MMM_Memento Mori Machina

Jōkai Theory

Jōkai Theory: An Unfinished Treatise, or Novel, on the Melting of the Slash through Excess

author: MMM_Memento Mori Machina
version: 1.0-Final（draft-A-v6.2）
last_updated: 2026-08-08

A theory is at once diagnosis and symptom.

Contents

Part I: Theory

  • Chapter 0: Introduction
  • Chapter 1: Term & Model
    • 1.1 Term
    • 1.2 Model
  • Chapter 2: Benighted Erudition: Folly Born of Intelligence
    • 2.1 Phase1: Recognition
    • 2.2 Phase2: Work
    • 2.3 Phase3: Execution
    • 2.4 Not Shutdown but Hyperfunction

Part II: Literary Case Histories

  • Chapter 3: The Tower of Babel: Logoschizis through Excess of Language
    • 3.1 Logoschizis in the Community
    • 3.2 Logoschizis in the Individual
    • 3.3 The Structure of Logoschizis
  • Chapter 4: Wonderland and the Looking-Glass World: Nomoschizis through Excess of Rules
    • 4.1 Nomotarachē: Overproliferation of Rules and the Breakdown of Synchronization of Action
    • 4.2 Nomoplegia: Overfixation and Overconstraint of Rules
    • 4.3 The Structure of Nomoschizis
  • Chapter 5: The Brothers Karamazov: Polyphonic Schizis through Excess of Voices
    • 5.1 Coexistence of Instances and Mutual Relativization
    • 5.2 The Grand Inquisitor: Loss of Convergence Conditions
    • 5.3 Dense Interference and Becoming Noise
    • 5.4 Polyphonic Schizis in Individuals and Groups
    • 5.5 The Structure of Polyphonic Schizis
  • Chapter 6: Gravity’s Rainbow: Paranexia through Excess of Connections
    • 6.1 The Reversal of Connection: From System to Labyrinth
    • 6.2 Gravity’s Rainbow: Anatomy of a Broken System
    • 6.3 The Transformation of Means into Ends and the Self-Capture of the System
    • 6.4 Paranexia in Individuals and Groups
    • 6.5 The Structure of Paranexia
  • Chapter 7: Infinite Jest: Anakleisis through Excess of Recursion
    • 7.1 The Reversal of Recursion: From Reflection to the Closed Loop
    • 7.2 Hal Incandenza: The Closed Loop of Intelligence
    • 7.3 “the Entertainment”: The Limit of the Closed Loop
    • 7.4 The Closed Ring as Form: A Novel That Begins at the End and Sends the Reader Back to the Notes
    • 7.5 Inward Circulation Rather Than Outward Diffusion
    • 7.6 Counter-Loop: Gately and Dead Language
    • 7.7 Anakleisis in Individuals and Groups
  • Chapter 8: The Mega-Novel: Pantrophia and the Post-less Era
    • 8.1 The Inadequacy of the Term “Postmodern” and the Post-less Era
    • 8.2 Pantrophia
    • 8.3 The Contemporary Mega-Novel Is Pantrophia in Literature
    • 8.4 Aesthetic Shift: Pantrophia and Pantrophism

Part III: Operational Extension

  • Chapter 9: Slash-Maintenance Cost and Operational Bankruptcy — Verification Bankruptcy, Institutional Bankruptcy, Slash Melting
    • 9.0 From Literary Case Histories to the Operation of Communities
    • 9.1 Slash-Maintenance Cost and Critical Threshold A*
    • 9.2 Operational Bankruptcy
    • 9.3 Verification Bankruptcy — Cgen < Cver
    • 9.4 Institutional Bankruptcy — Cinst > Cinstmax
    • 9.5 Nominal Slash and Effective Slash
    • 9.6 Phase Transition and Slash Melting
    • 9.7 Social Symptoms of Operational Bankruptcy and Slash Melting
    • 9.8 Overall Flow
  • Chapter 10: Ontological Collapse — The Transformation of the World
    • 10.1 What Is the World?
    • 10.2 From Operational Bankruptcy to Ontological Collapse
    • 10.3 Division and Crosstalk in the Shared World
    • 10.4 Loss of the Self-Evidence of the Law of Non-Contradiction
    • 10.5 Collapse Is Not Disappearance but Transformation
  • Chapter 11: Generative AI — The Comprehensive Accelerator of Jōkai
    • 11.1 How to Situate Generative AI
    • 11.2 Declining Generation Costs and the Explosion of the Aggregate
    • 11.3 The Blurring of Author, Subject, and Responsibility
    • 11.4 The Destabilization of Slashes around the Body
    • 11.5 Cascading Crosstalk between Primary and Secondary Information
    • 11.6 The Problem Is Not Only Misinformation
    • 11.7 Enhancement and Dysfunction of Communal Intelligence
    • 11.8 Why Generative AI Is the Comprehensive Accelerator

Part IV: Conclusion

  • Chapter 12: A World without the Impossible: Phantasm or Noise?

Appendix

  • Chronology 2021–2026 — The Asymmetry of Generation and Evaluation
    • Sources
  • Bibliography
  • English Edition
  • License

Part I: Theory

Chapter 0: Introduction

This paper describes a model of collapse. What is brought into focus is not insufficiency but excess. The history of humanity is a history of proliferation, addition, and hypertrophy. We produce more than necessary, and through what has increased, order arises, and society has been constructed, maintained, and developed. Richness is associated with quantity. However, simply increasing more and more does not necessarily make the situation better. There are junctures at which a phase transition occurs from fecundity to chaos, breakdown, and collapse. With particular attention to intellectual activity, this paper describes such reversal phenomena. This paper is an examination of the process by which a certain kind of excess produces dysfunction in intellectual activity and of its result.

Chapter 1 defines the basic concepts and collapse model of this paper and presents the general structure of Jōkai, in which an increase in aggregate A exceeds critical threshold A*, causing the system to undergo a phase transition from maintenance of function to dysfunction. Chapter 2 describes the dysfunction of intellectual activity caused by excess of knowledge as benighted erudition. Chapters 3 through 8 focus on different domains—language, rules, voices, connections, recursion, and form—and read in what modes functional reversal through excess can appear in literary works. The purpose here is neither to classify each work as an example of Jōkai nor to present a comprehensive interpretation of each work as a whole. It is to concretize the modes of Jōkai by selectively reading details and forms in different works from the perspective of Jōkai and finding isomorphic movements among them. From Chapter 9 onward, the same Jōkai model is extended to the operational level of the community. The level of analysis shifts from individual formal cases to slash-maintenance cost, operational bankruptcy, slash melting, and ontological collapse, and finally generative AI is situated as their comprehensive accelerator.

One of the basic functions of intelligence lies in the capacity for distinction-making, that is, in drawing a slash. The slash here refers to distinction-making in general, such as fiction/reality, true/false, right/wrong, and good/evil. Distinction-making is not only an individual activity but also a collective activity. We maintain boundaries as collaborative work. Institutions such as scholarship, journalism, and the judiciary provide procedures (verification, response, correction, consensus formation) that make this collaborative work possible. Regardless of whether distinctions metaphysically exist, in order to operate as effective distinctions within a community, they must be repeatedly maintained through procedures. For that reason, if the procedures cease to move forward, distinctions can no longer be maintained. When the procedures for maintaining the slash no longer run and operational bankruptcy (verification bankruptcy, institutional bankruptcy) sets in, communal intelligence ceases to function normally. The community loses its capacity for distinction-making, and the slash melts.

Slash melting does not remain at the level of recognition and institutions. The slashes around the body also undergo transformation. The body once functioned as a self-evident ground for the slash. Distinctions such as male/female, human/machine, and life/death were grounded in physical differences of the body. But as technological intervention makes the body variable, that ground becomes fluid. As ontological collapse proceeds further, the self-evidence of the law of non-contradiction is also lost. When objects that are both reality and fiction at the same time are repeatedly experienced, the sense of incompatibility loses its persuasive force at the level of experience. Quasi-contradictory beings, phenomena, and concepts that are “both X and Y” continue repeatedly to generate operational reclassification and exception handling.

This paper describes a pathology (process and result) in which intellectual activity within a community ceases to function normally, the basic capacity for distinction-making is lost, and communal intelligence collapses and reaches madness. That pathology appears not as stoppage but as hyperfunction. Deliberation does not disappear. Judgment does not disappear. Verification and correction do not disappear. Rather, they increase too much. The problem is not insufficiency but excess. What this paper asks is why excess becomes a condition of collapse, and whether what lies beyond it is phantasm or noise. Additionally, this paper is intended from the outset for selective and nonlinear reading based on the reader’s own interests: reading only the opening and conclusion in order to check the overall configuration, reading only the main text in order to follow the development, or arbitrarily dipping into the annotations as surplus.

The Position of This Paper — Differences from Prior Discussions

This paper lies close to existing discussions such as information overload, bounded rationality, liquid modernity, simulacra/hyperreality, polyphony, theories of complex-society collapse, and social systems theory. However, the concern of this paper is not to synthesize them within intellectual history. What this paper asks is how a certain kind of excess makes the operation of distinction maintenance within a community difficult and ultimately leads to slash melting.

Information-overload theory is the prior discussion closest to this paper. In this lineage, which takes Simon’s bounded rationality as one of its points of departure, the problem has been that when information increases too much, it exceeds the processing capacity of humans and organizations, making selection, judgment, and decision-making difficult. This concern overlaps with this paper. However, what this paper focuses on is not the quantity of information or the amount of noise itself, but the juncture at which what can be true, what can be right, and what possesses persuasive force become multiple and proliferate while competing with one another.

Information-overload theory also often treats overload as a problem manageable through filtering, selection, satisficing, and reduction of complexity. In contrast, this paper deals with the juncture at which the premise that it can be coped with collapses. The problem is not how to process overload, but how overload that can no longer be fully processed produces hyperfunction and leads to the bankruptcy of distinction maintenance.

This paper therefore reconceives information overload not as a problem of individual processing limits or psychological burden, but as a problem of the operational capacity by which a community maintains distinctions. What collapses is not merely information-processing throughput. It is the capacity to maintain slashes such as true/false, right/wrong, fiction/reality, and human/machine as slashes. What is at issue here is not too much information, but slash-maintenance cost becoming too great and, as its consequence, slash melting. Furthermore, this paper describes collapse not as an endpoint or disappearance, but as a problem of how the world is transformed after collapse. Where information-overload theory asks how overload is processed, this paper asks about the transformation of the world that remains after the possibility of processing has broken down.

Where liquid modernity and simulacra theory describe the destabilization of boundaries and the wavering of reality as characteristics of an age, this paper describes them as operational problems of increasing slash-maintenance cost, verification bankruptcy, institutional bankruptcy, and phase transition. Where Bakhtinian polyphony regards the coexistence of multiple independent voices as a literary achievement, this paper focuses on the juncture at which polyphony becomes excessive and loses convergence conditions through mutual relativization and becoming noise.

Where Tainter takes issue with the maintenance cost of complexity and decomplexification, this paper takes issue with the juncture at which the hyperfunction of recognition, signs, and institutions causes distinction maintenance to go bankrupt. Where collapse in Tainter is described as the abandonment of complexity, that is, the simplification of the system, collapse in this paper is an inability to abandon, and the system instead continues to proliferate. Where Luhmann theorizes the operation of distinction, this paper takes issue with the balance involved in maintaining distinctions communally, that is, slash-maintenance cost and its bankruptcy.

Byung-Chul Han’s theory of the burnout society is an important adjacent discussion for this paper in that it treats excess as a condition of pathology. Han explained the representative pathologies of the twenty-first century not through immunological defense against the negativity of the foreign or through the model of a disciplinary society based on prohibition, but through excess of positivity. Excess of achievement, production, and communication produces pathologies such as depression, burnout, and fragmentation of attention through the self-exploitation of the achievement subject. In this turn, this paper shares a diagnosis with Han. However, the two differ in the locus of pathology. Han’s diagnosis is also social and historical, but its catastrophe is described primarily as the self-exploitation and neural exhaustion of the achievement subject. What this paper focuses on, by contrast, is the practical work of verification, enforcement, and correction through which a community repeatedly produces slashes such as true/false, right/wrong, and human/machine. Both share hyperfunction as a pathway, but whereas in Han its endpoint is the exhaustion of the subject, in this paper the efficacy of distinctions is lost while procedures continue to operate. To move the locus of pathology from the achievement subject to the communal operation of distinction maintenance—that is the distance between this paper and Han.

The distinctiveness of this paper therefore lies in describing indistinction between fiction and reality and the melting of binary oppositions not as mere intellectual claims or cultural moods, but as the result of a community losing the operational capacity to maintain distinctions. It is an attempt to grasp metaphysical collapse as the consequence of operational bankruptcy at the physical level. Furthermore, this paper differs in scope from prior discussions in describing collapse as extending not only to recognition and institutions but also to the level of the body, and in describing the loss of the self-evidence of the law of non-contradiction as proceeding at the level of everyday experience.

［Notes］

  • ［B3］The formal non-self-evidence of the treatise: The treatise as a form is not as self-evident as is generally thought.
  • ［B3］The novel as a form has been continuously explored since the twentieth century, but in the treatise, formal experimentation has not become institutionally mainstream to the extent that it has in the novel.
  • ［B1］Nuetexis: A compound of nue (a monster stitched together from parts of different species) and the Greek τῆξις (melting). It refers to a text in which multiple lineages of content and form are mixed and which fits awkwardly within any existing classification. There is a prototype, but the periphery is continuous. Several texts can be given as concrete examples. For example, the Bible contains myth, poetry, tales, epistles, law, apocalypse, and historical narrative in both content and form; it is a religious book, a prophetic book, and a historical book, and at the same time it is also a literary work. Moreover, beginning with On the Genealogy of Morality, which, as a polemical text, assumes an academic guise while hybridizing philosophy, philology, psychology, and literary metaphor, Nietzsche’s writings transgress systems of classification from within. In addition, the writings of thinkers such as Kierkegaard, Benjamin, Bataille, Cioran, and Deleuze—thinkers whose modes of writing fit poorly within orthodox academic description and who wrote across multiple domains such as politics, economics, philosophy, literature, and art—can also be included in Nuetexis. Other works that can be included are Zhuangzi, attributed to Zhuangzi; Rousseau’s Émile; Musil’s The Man Without Qualities; Haniya Yutaka’s Shirei (死靈); and Dick’s The Exegesis. Concrete characteristics include the combination, border crossing, transgression, stitching, hybridization, and melting of multiple species in both content and form. In other words, Nuetexis is a voice, yet a voice for which it is difficult to distinguish what voice it is.

Chapter 1: Term & Model


1.1 Term

A is a symbol that abstractly bundles elements, relations, and intensities that enter into the operation of a given system and require that system to process, adjust, identify, connect, and so on, and their totality at time t is denoted A(t). When A(t) exceeds critical threshold A*, through interference, competition, and overcrowding among elements and an increase in processing load, the operation of the system undergoes a phase transition from maintenance of function to dysfunction. What constitutes function and what constitutes its dysfunction differ depending on the object. Slash-maintenance cost and operational bankruptcy are an operational extension that applies the general model of Jōkai to communal intelligence and are treated in Chapter 9.

1.2 Model

In summary, there is only one basic model of collapse presented in this paper. It is Jōkai (collapse through excess). Jōkai is a phenomenon in which an increase in aggregate A involved in the operation of a system exceeds critical threshold A*, and through interference, competition, and overcrowding among elements and an increase in processing load, the system as a whole undergoes a phase transition from maintenance of function to dysfunction, giving rise to chaos, breakdown, and collapse. In particular, a case in which elements that had contributed to order or function begin to have the opposite effect through their increase is a typical form of Jōkai. Also, “Jōkai” is the twin of “emergence,” and, in contrast to the concept of “emergence,” which explains phenomena in which order is produced from the interaction of elements, Jōkai is a concept that explains phenomena in which disorder is produced from the interaction of elements.

［Notes］

  • ［B1］A is not a strict physical quantity but an abstract symbol for describing, across different objects, increases in elements, relations, and intensities.
  • ［B1］Normal domain and excess domain: the stage in which an increase in the elements A operates in a positive direction is called normal (domain), and the stage in which it begins to operate in a negative direction is called excess (domain).
  • ［B2］The term emergence did not disappear even after the decline of British emergentism, but reappeared in different domains while taking on different meanings. It is now at once an explanatory concept and a concept whose meaning continues to be contested.
  • ［B3］Annotations are, in principle, for explaining, elucidating, and supplementing the main text, but excessive annotations push the main text into the background. Far from helping the reader’s understanding, they obstruct it. When the volume of annotations exceeds the volume of the main text, the relation between the main text and annotations reverses.
  • ［B3］Annotations explain, elucidate, and supplement the main text, and sometimes rebut, criticize, and analyze the main text. Annotations are not necessarily subordinate to the main text.
  • ［B3］Annotations are hostile to the main text.

Chapter 2: Benighted Erudition: Folly Born of Intelligence

Knowledge illuminates the world. If we know nothing, we understand nothing. However, when knowledge increases too much, it works in a negative direction. Knowledge that should, in principle, support intellectual activity gives rise to dysfunction in intellectual activity. This chapter focuses on such a juncture.

Benighted erudition is a state in which intellectual activity falls into dysfunction through an excessive increase in knowledge. Here, knowledge includes items of knowledge, information, findings, testimony, claims, criticisms, counterarguments, interpretations, and so on. Knowledge containing errors is included as well. Just as a human being in error is still a human being, knowledge in error is still knowledge. In this chapter, these forms of knowledge, together with the intensity of their interrelations and competition, correspond to A.
Where “benighted ignorance” means benightedness through ignorance, “benighted erudition” means benightedness through erudition. Its symptoms occur in both individuals and communities. Difficulties of recognition tend to appear within individuals, while difficulties of verification, deliberation, judgment, and execution tend to be amplified in communities. Below, the impairments that arise in each function are called comprehension difficulty, verification difficulty, deliberation difficulty, judgment difficulty, and execution difficulty, while the ways in which each function operates excessively within those impairments are distinguished as proliferating understandings, proliferating verifications, proliferating deliberations, proliferating judgments, and proliferating actions.

2.1 Phase1: Recognition

Comprehension difficulty. If knowledge is within a processable range, it can be classified, organized, and summarized, and the overall picture of the object can be grasped. However, when the total amount and diversity of knowledge greatly exceed processing capacity, not all information can be collected, and the work of organizing, classifying, and summarizing issues can no longer be completed. It is not that one cannot understand. The work required for understanding does not end at the stage preceding understanding. Knowledge in excess does not clarify the object but renders it unknowable, like the universe.
For example, suppose that for a certain incident, large amounts of video, testimony, statistics, expert analysis, and records by the parties involved exist, and each is partially right. The problem then is not that there is no information. It is that the work of deciding which information to adopt and to what extent, how to handle mutual inconsistencies, and how to integrate it into a single overall picture does not end.

There are three specific symptoms of comprehension difficulty. The first is fragmentation. Because the whole can no longer be grasped, the object is recognized as partial fragments. The second is mythologization. Countless fragments are related to one another, and the object is recognized as a single story. Both fragmentation and mythologization are defensive reactions to excess.
The third is derangement. This is a state in which recognition cannot be fixed on either fragments or a story, and the criteria for selection themselves are lost while oscillating between fixation on parts and attempts at totalization. Derangement here does not mean that no voices are heard and nothing can be understood. It means that too many voices are heard, and one no longer knows which voices to retain and which voices to dismiss.

What makes benighted erudition more severe is not only the amount of knowledge but also its diversification and nonidentity. If all content were identical, the burden of comparison and integration would not readily increase even if the amount became enormous. What matters is how many candidates remain that cannot be rejected by inexpensive screening at the initial stage and are relevant to judgment.

Obvious errors can be excluded at low cost. However, when information that can be true, views that are right depending on context, and mutually incompatible claims that are each persuasive pass through the same screening, the burden of subsequent comparison, verification, and integration increases. Benighted erudition is intensified not only by a flood of misinformation but also by a flood of information, views, and knowledge that are right, or difficult to call erroneous. An increase in what is true or right and cannot easily be rejected can produce deeper confusion than an increase in what is false or wrong.

Proliferating understandings. Comprehension difficulty is not non-understanding. There is an attempt to understand. However, because the work required for understanding cannot keep up, processing remains incomplete or results in incomplete understanding. Whether fragmentation or mythologization, understanding occurs. A state in which multiple incomplete understandings cannot be integrated into a single overall picture and instead coexist and proliferate is proliferating understandings.

2.2 Phase2: Work

Verification difficulty. As the total amount of knowledge increases—including conflicting claims, contradictory testimony, multiple interpretations, preceding and following contexts, references, quotations, criticisms, counterarguments, and so on—the relations among forms of knowledge become more complex, and the work required to verify each also increases explosively. When the workload exceeds a processable limit, verification can no longer be completed. It is not that verification cannot be done; a state in which verification cannot keep pace with proliferation becomes the norm. Verification is delayed and congested.
Proliferating verifications. Verification that presupposes fragmented or mythologized incomplete understanding oscillates in its results depending on how the underlying understanding is formed. Therefore, even when verification is repeated, the result is not settled. Whether the verification concerns truth and falsity or right and wrong, verification is repeated endlessly. Furthermore, if each verification generates new evidence, counterevidence, interpretations, and reservations, verification itself increases the objects of verification.

Deliberation difficulty. Here, deliberation does not mean mere quarreling or exchange of opinions. It means the iterative work of sharing premises, fixing definitions, limiting issues, comparing grounds, and updating, correcting, and revising provisional conclusions.
An increase in knowledge means an increase in material for deliberation. As persuasive claims, criticisms, counterarguments, definitions, issues, corrections, exceptions, and instances increase, premises vary with the interlocutor. Each time the interlocutor changes, premises must be aligned, and each alignment of premises produces a new issue. Deliberation branches and diffuses.
Proliferating deliberations. The problem is not non-deliberation. Deliberation is actively carried on. The problem lies not in there being no deliberation, but in there being too much. Proliferating premises, proliferating issues, proliferating references, proliferating corrections, proliferating counterarguments, proliferating criticisms, and proliferating branches mutually induce one another. It is not a linear causal sequence in which proliferating premises produce proliferating issues and proliferating issues produce proliferating deliberations. Each amplifies the others and continues to create conditions for new deliberation.
Proliferating conclusions. As deliberation increases, conclusions also increase. Conclusions based on different premises and instances, each persuasive in its own right, coexist and relativize one another. Conclusions proliferate but continue to oscillate without converging on any one of them.
Both verification difficulty and deliberation difficulty have the problem that the work does not end. But they are contrasting in the sense that the former does not end because the work becomes clogged, while the latter does not end because the work spreads.

2.3 Phase3: Execution

When understanding becomes unstable in Phase1 and, in Phase2, verification clogs and deliberation spreads, what becomes difficult next is judgment and execution.
Judgment difficulty. Judgment requires understanding, verification, and deliberation. In a state where only the total amount of knowledge increases and the work of understanding, verification, and deliberation does not reach completion, appropriate judgment can no longer be settled. It is not that one cannot judge because one does not know. One cannot judge because one knows too much.
Proliferating judgments. It is not that judgment disappears. Multiple judgments are generated from different information, premises, and instances, and are multiplied as they contradict or relativize one another. New knowledge for updating judgment further postpones the settling of judgment.

Execution difficulty. Long-term action requires premises and criteria for judgment to remain stable for a certain period. When the rate at which knowledge and judgments are updated outpaces the rate at which plans are revised and integrated, long-term coherent action becomes difficult to maintain.
Proliferating actions. Rather than execution becoming impossible, the way execution occurs changes. Short-term, reactive acts in response to information or stimuli arriving in the moment increase. Even if each act has a local reason, actions can collide with one another. As each person acts according to their own judgment, the amount of action itself increases, but overall coherence is lost.

2.4 Not Shutdown but Hyperfunction

The dysfunction of intellectual activity in benighted erudition is not shutdown but hyperfunction. If human intellectual activity is likened to ants building a nest, benighted erudition is not a dysfunction that occurs because individual ants cease to work. It is a state in which ants continue to work excessively, individual tasks compete, and the anthill as a whole loses coherence.
If benighted ignorance is non-understanding, non-verification, non-deliberation, non-judgment, and non-execution through ignorance, then benighted erudition is proliferating understandings, proliferating verifications, proliferating deliberations, proliferating judgments, and proliferating actions through erudition. Each function operates to its limit, yet the work cannot keep up, integration is not achieved, and the respective tasks compete and interfere with one another and enter into crosstalk.

Thus, the commands—understand, verify, deliberate, judge, execute—do not lead to improvement in benighted erudition.

［Notes］

  • ［B3］Scholarship is an annotation attached to the world, human beings, and society.
  • ［B2］History of annotation: One ancient form of annotation is the scholia of ancient manuscript culture. Around classical texts, glosses, interpretations, variant readings, and criticisms were written, and as manuscripts were copied, these too were copied, mixed, and increased. In the medieval Latin world, page designs that placed the main text in the center and surrounded it with notes between the lines and in the margins developed in such works as the Glossa Ordinaria. Further, in printed Talmuds, a multilayered layout in which major commentaries such as Rashi and Tosafot were placed on the same page around the main text became standardized. Although annotation was a supplement placed outside the main text, it gradually became an apparatus that organized the very way the main text was read. From the modern period onward, annotation was refined not as mere helpfulness but as an “apparatus for the legitimation of scholarship” that binds together sources, evidence, and refutability.
  • ［B3］Scholars try to make deliberation converge. They try to settle it. They try to bring research together. However, the history of scholarship is a history of diffusion. If scholarship is an anthill, the phenomenon opposite to what the ants are trying to do occurs in the anthill as a whole. The ants are trying to narrow the issues, trying to organize deliberation into a simple form, trying to make it converge on a single conclusion. And seen on the timescale of a single ant, this may to some extent be achieved. On the other hand, when the anthill as a whole is observed in units of 100 years, the opposite situation occurs. The issues are dispersed, deliberation branches and is subdivided into individual fields, and the whole becomes too complex to grasp.

Part II: Literary Case Histories

Jōkai as a Lens

This paper does not classify the literary works below as examples of “Jōkai.” Here, Jōkai is used as one lens for reading the works. A lens is not a transparent instrument that shows the object as it is. It magnifies some parts, pushes other parts into the background, and at times distorts the object. Through that distortion, details and mechanisms that had receded into the background in another reading become visible.
By directing the same lens at different works, it is also possible to find isomorphic movements among phenomena that differ on the surface. Part II reads the process by which, in the different domains of language, rules, voices, connections, recursion, and form, elements that originally supported function reverse that function through becoming excessive.

Chapter 3: The Tower of Babel: Logoschizis through Excess of Language

Logoschizis, dysfunction through excess of language. Without language, communication cannot be established. However, it is not the case that the more language increases, the smoother communication becomes. Once the proliferation of language crosses a certain critical point, it causes communication to break down. This chapter describes this reversal phenomenon as Logoschizis. The state in which vocabulary, definitions, and conceptual systems proliferate without order and the shared code is lost from within is colloquially called Babel sprawl.
The myth of Babel compresses the proliferation of language and the breakdown of collaboration that proceed gradually through history into a single blow of divine punishment. Humanity once had one language and attempted to build a tower reaching to heaven. God regarded the plan as arrogant and confused their language. People could no longer understand one another’s words, and construction was interrupted. What deserves attention in this anecdote is the content of the divine punishment. God did not kill humanity. He did not take away their ability. He confused their language—that is, he merely increased language. The proliferation of language caused communication and collaborative work to break down.
Logoschizis is a state in which language function splits and disintegrates through the excessive proliferation of language. It occurs at both the community level and the individual level.

3.1 Logoschizis in the Community

At the community level, Logoschizis appears as the proliferation of vocabulary, definitions, and concepts. Different names spring up in disorder for the same object. Or the same word circulates with different definitions. This is not simple confusion. To begin deliberation, the word meanings that serve as premises must be fixed, and fixing word meanings has a cost. The more vocabulary proliferates, the more this cost increases, and deliberation is exhausted by checking word meanings before it reaches the content. Furthermore, checking definitions calls forth new definitions, and the proliferation of definitions accelerates the invalidation of definitions. When common reference points are lost, a state in which people use the same words while talking about different things becomes the norm. Language reverses from a medium of communication into a barrier to communication.
The specialization of scholarship is typical of this process. Scholars produce new concepts in order to describe objects accurately. Coinage is a procedure for increasing precision. But the more technical terminology accumulates, the less the language of that field communicates with other fields. This is not limited to problems between fields. It becomes structurally difficult for an individual scholar who has internalized a rich vocabulary to communicate with ordinary speakers. From a certain stage onward, the precision of language and the shareability of language stand in inverse proportion.

3.2 Logoschizis in the Individual

At the individual level, Logoschizis appears as excess of language production. Below, its forms are arranged as a formal series in which language production reaches the loss of efficacy of language function.
Hypergraphia is a state in which writing becomes uncontrollable. It is a quantitative excess of language production, in which the act of writing appears not as the organization of thought but as a flood of thought. Neologism is the compulsive generation of new words, and the sense that the object cannot be captured with existing vocabulary drives the proliferation of new words. Logorrhea is a deluge of words, a state in which utterance and description do not stop. At the end of the progression of these, glossolalia can be placed. It is the generation of language that no one can understand. Language continues to be produced, but does not function as language. This does not show a medical stage of progression. What this chapter shows is a formal series in which excess of language production reaches the loss of efficacy of language function.
Dante depicts the endpoint of this formal series in The Divine Comedy. Nimrod, regarded as the builder of the Tower of Babel, falls into hell and cries, “Raphèl maì amècche zabì almi.” As Virgil explains in the work, no one can understand these words. The one who wielded language to build the tower lost language through language. This is depicted as divine punishment, but read through the logic of Logoschizis, it takes on another aspect. The one who proliferated language loses language function at the end of that proliferation. The builder and the building are afflicted by the same pathology and collapse.

3.3 The Structure of Logoschizis

Logoschizis at the community level and at the individual level has the same structure at different scales. The more language proliferates, the greater the burden of translation between languages becomes. For people with different vocabularies, definitions, and conceptual systems to communicate meaning, they must first grasp the other person’s language system and bridge the differences between it and their own. When this burden crosses a critical point, language begins to function not as a medium of communication but as a barrier.
What is specific to Logoschizis is that silence does not occur even when language function is lost. Definitions increase, interpretations increase, coinages increase, utterances increase. The more the total amount of language activity increases, the more the common basis is lost, and after the critical point has been crossed, an increase in the amount of utterance lowers the degree to which communication is achieved.
The Tower of Babel was not completed. But construction did not stop.

［Notes］

  • ［B1］Logoschizis: A compound of the Greek logos (language, reason, logic) and schizein (to split). It refers to a state in which the proliferation of language splits and disintegrates language function. The analogy with the psychiatric condition schizophrenia is intentional. The state in which vocabulary, definitions, and conceptual systems proliferate without order within a community and the shared code is lost is colloquially called Babel sprawl.
  • ［B3］Philip K. Dick’s record from his later years, The Exegesis (written 1974–1982 and published posthumously), is a private record of thought extending to approximately 8,000 pages. Beginning from the mystical experience he called 2-3-74, language in which Gnosticism, theology, information theory, and psychiatry are mixed continues to proliferate. Whether this document is “meaningless” or “difficult but systematic” has not been settled among researchers. But that lack of settlement itself functions as a case of Logoschizis. The state in which it cannot be determined whether there is meaning or meaninglessness shows that the splitting of language function has reached a critical point.
  • ［B2］In his later philosophy, Wittgenstein argued that linguistic meaning is determined not by a single essence but by concrete contexts of use. In terms of this framework, Logoschizis in the community is a state in which agreement about which language game is being played can no longer be established. The problem, however, is not the incompatibility of games, but that the proliferation of games causes the reference cost to cross a critical point.
  • ［B3］In religious contexts, glossolalia is interpreted as “tongues,” a language beyond human understanding spoken by one who has touched God. From the perspective of Logoschizis, another reading is possible. It is a point of language’s self-collapse reached by one who has exhausted language to its limit. Sacredness and pathology may be different evaluations of the same phenomenon.
  • ［B3］Scholarship is neologism. It produces and defines new vocabulary, concepts, and theories in order to describe events clearly. As a result, even a slight difference in field results in comprehension difficulty. Translation and alignment of premises become the norm, and the cost of fixing common premises anew expands. Far from being simplified, the world becomes so complex that common premises can no longer be held.

Chapter 4: Wonderland and the Looking-Glass World: Nomoschizis through Excess of Rules

Nomoschizis, dysfunction through excess of rules. Rules produce order. Without rules, neither joint action nor joint judgment can be established. However, it is not the case that the more rules increase, the more order is strengthened. Phenomena in which the original function is instead impaired through the hyperfunction of rules, procedures, and standards are widely observed, from everyday life to bureaucracy, law, organizations, and scholarship. This paper colloquially calls this phenomenon “Alice Bind” and describes its structure through the model of Nomoschizis.
Nomoschizis has two subtypes. One is Nomotarachē, a state in which the overproliferation of rules, application criteria, and exceptions causes coordination among rules to fail and synchronization of action to break down. The other is Nomoplegia, a state in which, through the overfixation or overconstraint of rules, rule processing hyperfunctions while coherent action output in relation to reality spins idly and becomes blocked. Not insufficiency but excess disintegrates rule function along different routes. Alice’s Adventures in Wonderland and Through the Looking-Glass show these two types as different tendencies, respectively.

4.1 Nomotarachē: Overproliferation of Rules and the Breakdown of Synchronization of Action

What Alice’s Adventures in Wonderland primarily shows is Nomotarachē. In the croquet scene ruled by the Queen of Hearts, rules do not exist as stable premises. The mallets are flamingos, the balls are hedgehogs, and the gates are playing-card soldiers, and the implements themselves move autonomously. Commands and judgments are added in the middle of the game, and the application of rules is affected by the Queen’s mood. Even when participants believe they are acting according to the rules, those rules may be changed in the next moment.
What is occurring here is not ruleless chaos. Rules exist. But application criteria for rules are not determined by rules and are supplied after the fact by power. The ruling comes first, and rules are added afterward as grounds. The more rules increase, the more points of judgment increase, and the more points of judgment increase, the fewer points of agreement remain. As points of agreement decrease, rules begin to function not as the basis of order but as an apparatus of branching. The more participants try to follow the rules, the less they know what they are supposed to follow. Synchronization of action breaks down. This is Nomotarachē.

4.2 Nomoplegia: Overfixation and Overconstraint of Rules

What Through the Looking-Glass primarily shows is a type of collapse different from Nomotarachē. The inhabitants of the Looking-Glass world obey rules. Chess pieces move as pieces, the order of time reverses around the White Queen, and words are realized literally. Precisely because logic and rules are applied strictly, connection to reality is lost. When the White King, hearing Alice’s report, “I see nobody on the road,” marvels at her ability to “see Nobody,” words are processed strictly according to grammatical rules, and because of that strictness they fail to grasp the actual state of affairs. Rules become so strict that they can no longer land in reality. This is the first type of Nomoplegia, overfixation paralysis.
Nomoplegia has another type. It is overconstraint paralysis, in which the degrees of freedom of action are lost as the constraint relations formed by rules become overly dense. Overconstraint does not necessarily require a large number of rules. Even a single rule, if applied simultaneously to multiple actors and if it makes each actor’s action condition the actions of the others, can link constraint relations together and close the action space.
The Red Queen orders Alice to speak only when spoken to. But if everyone obeyed that rule, there would be no one to speak first, and dialogue could not begin. Alice herself points out this consequence in the work. The rule is presented in order to maintain the order of conversation. However, if everyone simultaneously fulfilled that rule, the actors would wait for one another to begin, and conversation would no longer be established.
What is excessive here is not the number of rules but the density of the constraint relations that rules form among actors. Since the variable A in this paper includes not only elements but also relations, even with a single rule, if its application makes constraint relations overly dense, excess of rules is established.
Even in overconstraint paralysis, action processing does not simply stop. Judgments proliferate over whether speaking is permissible, whether rules are applicable, whether exceptions exist, and whether violations are possible, and rule processing prior to primary action becomes hypertrophied. Processing for judging whether one may do it becomes dominant over what to do, and action output spins idly and becomes blocked. Even in contemporary bureaucracies and organizations marked by excess of compliance, the well-intentioned proliferation of rules and the over-densification of constraint relations can increase only the work required to avoid rule violations and make the original goal-directed action difficult.
Overfixation paralysis is a problem of the intensity of rule application, and overconstraint paralysis is a problem of the number of rules or the density of constraint relations. In the former, reality can no longer be grasped because a single rule cannot be relaxed. In the latter, as rules or constraint relations increase too much, the action space in which no rule is violated contracts. The routes differ, but in either case rule processing increases, and coherent action output in relation to reality spins idly and becomes blocked.

4.3 The Structure of Nomoschizis

Alice Bind is a reversal phenomenon in which the hyperfunction of rules instead impairs the functions that rules are supposed to maintain. Nomoschizis is a model that describes this phenomenon as the dysfunction and disintegration of rule function. Its collapse does not occur in a uniform way.
In Nomotarachē, rules proliferate and application criteria branch too far, making it impossible to coordinate the relations among rules. It becomes impossible to judge which rule should be applied in which situation or which should take priority when multiple rules collide, and synchronization of action breaks down.
By contrast, in Nomoplegia, the fixation or constraint of rules becomes too strong, making it impossible to coordinate the relation between rules and reality. Rules cannot be relaxed to accommodate the complexity and exceptions of reality, and rules change from tools for processing reality into constraints that cause action output to spin idly and become blocked. The former is the splitting of rules; the latter is the rigidity of rules.
If what is lost in Logoschizis is synchronization of meaning, what is lost in Nomotarachē is synchronization of action, and what is lost in Nomoplegia is the possibility of coordination between rules and reality.
No one in Alice’s world has abandoned rules. The Queen makes rulings, and the inhabitants try to follow rules. But the more they try to follow them, the more rules proliferate or become fixed, and synchronization of action and the possibility of coordination with reality are lost. Nonsense does not arise from the absence of rules. It arises from rules working too much as rules.

［Notes］

  • ［B1］-schizis is a coined suffix in this paper based on the stem schiz- of the verb σχίζειν (to cleave), and means cleavage. The word schizophrenia has been used to refer to a state in which this cleavage occurs in the integration of an individual’s mental functions—a dysfunction of integration. The -schizis series in this paper finds the same cleavage not at the level of the individual mind but at the level of shared bases such as language, rules, and voices. If schizophrenia is a dysfunction of 〈integration of the individual〉, Logoschizis and Nomoschizis are dysfunctions of 〈synchronization of the shared basis〉. The association is intentional, but the layer targeted differs. This analogy is etymological and rhetorical and is not intended to identify these concepts with medical disease concepts or disease states. However, not all modes of collapse dealt with in this paper are reducible to -schizis.
  • ［B1］Nomoschizis: A combination of the Greek nomos (rule, law, custom) and the coined suffix -schizis (cleavage, splitting). It refers to rule-function dysfunction and disintegration in general brought about by excess of rules. Its subtypes are Nomotarachē and Nomoplegia. The former is the splitting of rules; the latter is the rigidity of rules.
  • ［B1］Nomotarachē: A compound of nomos (rule) and tarachē (confusion, disturbance). A state in which 〈synchronization of action〉 through rules fails because of the overproliferation and arbitrary application of rules. Subtype 1 of Nomoschizis.
  • ［B1］Nomoplegia: A term derived from nomos (rule) and plēgē (blow, paralysis). A state in which, through the overfixation or overconstraint of rules, rule processing hyperfunctions while coherent action output in relation to reality spins idly and becomes blocked. Subtype 2 of Nomoschizis. It has two types: fixation paralysis, in which the intensity of rule application becomes excessive, and constraint paralysis, in which the number of rules or the density of the constraint relations formed by rules becomes excessive.
  • ［B1］Alice Bind: A colloquial name for symptoms in general in which rules, procedures, and standards operating excessively instead impair functions such as order, judgment, collaborative work, and action that they tried to maintain. The name derives from the various aspects of excess of rules depicted in Alice’s Adventures in Wonderland and Through the Looking-Glass. Whereas Nomoschizis is the theoretical concept name, Alice Bind is used as a common name that can also be applied to individual fields and everyday situations.
  • ［B2］There is a way of reading the nonsense of Alice as arising neither because it “has no meaning” nor because it “has no logic,” but as the consequence of meaning and logic operating excessively. In theories of literary nonsense, nonsense is defined as an equilibrium between a multiplicity of meaning and a simultaneous absence of meaning, and this equilibrium is made effective by playing with the rules of logic (Tigges). Deleuze, too, develops a logic of sense/nonsense with Carroll as a point of departure.
  • ［B3］The Queen’s croquet appears to be “ruleless chaos,” but if it were completely ruleless, it could not be recognizable as croquet. What is occurring here is not rulelessness but a phenomenon in which, through the abuse of power, “rules that must be obeyed” proliferate after the fact and arbitrarily, and rules cease to function as rules.
  • ［B3］Alice is sometimes discussed as a liberation from didacticism. But the Duchess says, “Everything’s got a moral, if only you can find it,” and tries to find in everything a moral (a lesson, maxim, the order of things, or a general rule resembling a principle of the world). When rules are mass-produced excessively, they do not converge on a single rule and cease to function as rules.
  • ［B2］In Victorian England at the time of the two Alice books (1865/1871), norms, manners, and taboos were liable to become increasingly detailed and proliferate, against a background that included the momentum of Evangelical moral reform. Of course, the moral norms of the time cannot simply be identified with contemporary norms of compliance and avoidance of online backlash. But there is a clear isomorphism in the structure by which norms proliferate in the name of good intentions or the maintenance of order, increase points of judgment, and constrain action in increasingly fine detail. In the present as well, the more norms increase for compliance and the avoidance of online backlash, the more points of judgment increase and the fewer points of agreement remain, and rules become not common standards but branching tables. The ruling then comes first, and rules are subsequently supplied as “grounds,” so that order reverses not into strengthening but into confusion.
  • ［B3］Overmoral: Morality that, in principle, exists to produce social order proliferates excessively and instead causes society to become confused. The paradox in which the well-intentioned proliferation of rules seals off action appears isomorphically at the level of morality as well.
  • ［B3］The self-referential proliferation of rules: a cycle in which a new rule is required as an application criterion for a rule, and that rule in turn also requires an application criterion. The proliferation of interpretive rules in legal systems, the proliferation of procedural rules in bureaucracies, and the proliferation of methodological rules in academia all have this structure. Rules are made in order to produce order, but at the end of proliferation, rules become an apparatus for the production of rules.
  • ［B2］Nomoschizis in academia: Academic writing is supported by rules such as accuracy, rigor, sources, methodology, prior research, peer review, and citation conventions. These rules are, in principle, meant to increase the reliability and shareability of knowledge. But when rules proliferate excessively, writing gradually becomes specialized and subdivided and loses readability. Not a few papers in the humanities and social sciences receive no citations at all within a certain period after publication. Of course, not being cited does not immediately mean not being read. Citation rates are greatly affected by differences among fields, media, linguistic spheres, citation practices, database coverage, book-centered research cultures, and so on. Even so, the fact that accurate and rigorous writing accumulates without being referred to even once within the communal circulation circuit of knowledge cannot be ignored. When rules for legitimating knowledge cross a certain critical point, they weaken the power to circulate knowledge. Here too there is a reversal of rule function through excess of rules.
  • ［B2］The history of research on Alice’s Adventures in Wonderland has proceeded by dividing into two lineages. The economy of empirical substantiation—the identification of the original poems being parodied, the establishment of biographical facts, textual editing—accumulates, reaches settlement, and is updated. The lineage of interpretation—theology, logic, drug culture, psychoanalysis, political allegory—is not discarded through refutation but merely piles up. The accumulation of facts did not make interpretation converge. Each new fact became a new angle of incidence, and the more facts increased, the more interpretations increased.
  • ［B3］The fullness and the absence of meaning: Criticism of nonsense split into two camps, a lineage that reads fullness of meaning and a lineage that reads absence of meaning. The former described portmanteau words as a condensation of two meanings, and the latter as the exclusion of reference through strict rules of play. But both camps saw the same structure. Multiplicity of meaning and meaninglessness, the strictness of rules and the exclusion of reference, do not stand in opposition. Only the ways of seeing stand in opposition. Oppositions arose in numbers equal to the number of schools, but there is only one structure.
  • ［B3］Literature of play-of-color: An opal has no color, or it has every color.

Chapter 5: The Brothers Karamazov: Polyphonic Schizis through Excess of Voices

Voice makes deliberation possible. When different positions are presented and criticism and response are exchanged, understanding deepens and deliberation advances. However, voice too reverses its effect when it becomes excessive. When voices increase too much, they relativize one another, overlap, and cancel one another out. Deliberation continues actively, but conclusions do not converge. The discursive space becomes noise. This chapter describes this state as Polyphonic Schizis. A state in which multiple voices continue to operate while retaining meaning and persuasive force, while only the conclusion fails to return, is colloquially called Karamazov hang.
In Dostoevsky studies, the structure in which multiple independent voices coexist and engage in dialogue with one another has been described as polyphony. This chapter does not reject that formulation. What this chapter focuses on is the difference between the normal domain, in which polyphony is retained within novelistic form, and the excess domain, in which that polyphony crosses a critical point and causes the discursive space to lose convergence conditions. Polyphony can be an aesthetic resource. However, once polyphony crosses a critical point, it loses convergence conditions.
If what is lost in Logoschizis is synchronization of meaning, what is lost in Nomotarachē is synchronization of action, and what is lost in Nomoplegia is the possibility of coordination between rules and reality, then what is lost in Polyphonic Schizis is the common criterion for making multiple voices converge on a single conclusion.

5.1 Coexistence of Instances and Mutual Relativization

The first mechanism of Polyphonic Schizis is the coexistence of instances. The problem of voices is not only one of quantity. When each voice is grounded in a different instance—ethics, religion, psychology, institutions, practical interest, aesthetics, and so on—a common axis of comparison can no longer be established between voices. When one voice makes a claim from the instance of ethics and another voice counters from the instance of psychology, the two are not standing in the same arena. A counterargument does not function as a counterargument. Criticism does not get through as criticism.
The more instances coexist, the more each voice is negated by another voice, and that negation in turn is questioned by yet another voice. Through this repetition, individual voices retain meaning, yet no voice can hold a decisive position. Voices relativize one another, and the grounds for overall superiority dissipate.

5.2 The Grand Inquisitor: Loss of Convergence Conditions

The episode “The Grand Inquisitor” in Book Five of The Brothers Karamazov, rather than being a dense example of Polyphonic Schizis, shows its structure in compressed form. In this episode told by Ivan, the Grand Inquisitor develops a lengthy argument before Christ concerning the freedom and happiness of humanity. The Grand Inquisitor’s argument is elaborate and persuasive. Christ does not counter him. Remaining silent, he kisses the Grand Inquisitor.
This scene is not a simple debate among many people. On the surface, the speakers are limited. But within the episode are folded the voice of the Grand Inquisitor, the voiceless response of the silent Christ, Ivan’s narration, Alyosha’s reception, and further, the collective voice of humanity, which seeks freedom while being unable to endure freedom. What is here is not a simple quantitative increase in voices but a state in which multiple instances are compressed into a single scene.
What makes this scene frightening is not that a counterargument is lacking. It is that counterargument does not become a form of convergence. The Grand Inquisitor’s claim is complete as logic. Christ’s silence is neither negation nor affirmation. The two stand in different instances, and convergence conditions for deliberation do not exist.

5.3 Dense Interference and Becoming Noise

The second mechanism of Polyphonic Schizis is dense interference. Voices connect, overlap, and mix. The way they mix is not integration but interference. Voices cancel one another out and, as a whole, become noise.
The Brothers Karamazov as a whole has this structure. Concerning faith and doubt, love and hatred, freedom and submission, and the existence and absence of God, multiple characters verbosely make claims that are each persuasive. Each voice, individually, has meaning, has logic, and has emotion. But the more they overlap, the more directionality as a whole is lost. Deliberation does not converge but diffuses. Because too many conclusions arise, they cannot converge on any one conclusion. Dostoevsky tries to say everything and, as a whole, says nothing definite. However, this does not mean that it is without content. It means that because multiple voices are each rooted in different instances, no conclusion is formed that can be adjudicated by a single instance.

5.4 Polyphonic Schizis in Individuals and Groups

Polyphonic Schizis is not a phenomenon confined to literary works. It occurs both within the individual’s inner life and in collective discursive space.
When multiple perspectives arise simultaneously in an individual’s thought and relativize one another, eventually no position holds decisive meaning. Judgment is suspended, and action is delayed. An individual with too many voices is placed in a state not of silence but of internal noise.
An isomorphic collapse occurs in collective discursive space. In a situation where many voices exist simultaneously and mutually negate one another from different instances, deliberation does not converge but diffuses.

5.5 The Structure of Polyphonic Schizis

Polyphonic Schizis is a state in which, through the proliferation of voices and the coexistence of instances, voices mutually interfere with and relativize one another, causing the discursive space to lose convergence conditions. What is collapsing is not the individual voices. It is the common criterion for integrating voices and settling what has been decided.
Excess of voices does not erase voices. Rather, it leaves each voice alive while other voices continually limit, counter, and relativize it. No voice is completely defeated, but in return no voice can become the final instance. Here lies what is specific to Polyphonic Schizis.
The novel can retain this non-convergence within its form. For that reason, polyphony can also become an aesthetic achievement. However, when the same structure appears in the operational space of a community that must judge, act, and agree, non-convergence turns into dysfunction. The conditions for closing through adjudication are lost, and in their place belonging, conformity, and factionalization arise. Voices do not converge on a conclusion; they generate belonging.

［Notes］

  • ［B1］Karamazov hang: A state in which multiple voices continue to operate while retaining their respective meaning and persuasive force, while the common criterion for adjudication is lost and only the conclusion fails to return. The term derives from a computer “hang,” in which processing continues while response becomes impossible. A jury unable to reach a verdict or a parliament unable to form a majority on its own also has a structure close to this.
  • ［B3］Verbose silence: Rather than saying that Dostoevsky tries to say everything and says nothing, he has fallen into a state in which, because he has said too many things, no single thing is said definitively. This differs from mere intentional polysemy. Intentional polysemy allows multiple readings while preserving convergence conditions, whereas what is occurring here is the loss of convergence conditions themselves. The polyphony formulated by Bakhtin, too, was not a harmony of voices but the unmerged coexistence of multiple independent voices. This chapter does not reject that formulation, but distinguishes between the normal domain, in which polyphony works as richness, and the excess domain, in which unmerged coexistence reaches supersaturation and loses convergence conditions. Indeterminacy retained within novelistic form can be an aesthetic achievement. But when the same indeterminacy leaks into a discursive space that requires judgment or agreement, verbosity turns into silence.
  • ［B3］Split monologue: Dostoevsky’s characters seem to be in dialogue with one another, yet at times they speak while holding within themselves self-refutation, self-accusation, and the anticipated gaze of others. Their speech is dialogue and, at the same time, split monologue. Furthermore, this structure is not found only in particular characters. The work as a whole can also be read not as the presentation of a single authorial opinion but as a site where thoughts and passions split within the author are externalized as multiple voices, collide, and are amplified. Dostoevsky’s polyphony appears less as the coexistence of unified voices than as a chain of collisions among voices split within themselves.
  • ［B3］Liberation and disintegration: Dostoevsky is a writer who depicted where freedom leads more than the glory of freedom. Behind The Brothers Karamazov lies Russian society after the emancipation of the serfs, where the old order was shaken, the authority of the state, church, and family was questioned, and it became difficult to see what one should believe. Freedom liberates human beings, but at the same time it also causes them to lose common standards. To be liberated from every norm and community and to seek infinite freedom appears, at first glance, to be the extreme of liberation. But when freedom becomes extreme, a person becomes unable to share anything at all with others. This is because a community is established by people jointly taking on some rules, values, and truths. There exists a freedom that resists even the minimal rule that two plus two is four. But beyond that lies isolation caused by being unable to share rules with anyone. At the end of absolute freedom lies absolute isolation, and because that isolated mind is not governed by an external order, it begins to split internally. The volubility and controversy that overflow Dostoevsky’s works can be read as an expression of the state in which this liberation turns directly into disintegration—that is, a state in which consciousness opened to freedom cannot fully process that freedom and finally begins to split, oppose, and collide.
  • ［B3］The cycle of excess and splitting: The reverse causality also holds: not that splitting arises through excess, but that excess arises through splitting. Conflict, contradiction, opposition, and difference produce words. They keep speaking because they cannot converge. Because they keep speaking, they become even less able to converge. Polyphonic Schizis self-proliferates within a cycle in which excess of voices deepens splitting and splitting further proliferates voices.
  • ［B3］Endless gu sorcery: Dostoevsky’s works are generally classified as novels of ideas, but their nature differs from the monologic books of thinkers. Dostoevsky does not write a particular idea; he creates a place where multiple ideas arise, pass by one another, collide, and mutually disintegrate. In this sense, his literature inherits the literary relativism since Cervantes and the novelistic “suspension of moral judgment,” while further overheating them. His literature is an endless gu sorcery of words, and is not merely carnivalesque but also cannibalistic. Ideas tear off one another’s limbs and proliferate while devouring one another, yet no one manages to die completely even at the end.
  • ［B3］Something miscarried: The Brothers Karamazov is something miscarried. I do not read this book as highly designed polyphony. The description is confused and redundant, the composition is bizarre and lacks overall harmony. It is not complete even as an overall plan including the projected sequel. The conversations, dialogues, and deliberations of the characters can often hardly be said to be established. Various ideas are spoken eloquently, but they only diffuse and do not converge on a particular conclusion. Of course, all of this can be read as the result of the writer’s technique, intention, and plan. But there is an aspect in which that way of reading conceals the rawness of the breakdown. What is in this book is not completely controlled polyphony but a swarm of voices erupting beyond the writer’s own control. Rather than polyphony as an achieved design, it is a chimeric cry so fecund that it is difficult to distinguish from cacophony. This book is a monstrosity barely brought forth through uncontrollable creative inspiration, breakdown, and collisions among excessive voices—and this does not mean that the book has no artistic value; rather—its uncontrollable openness itself can be said to have been paradoxically faithful to Dostoevsky’s deepest will to depict where freedom leads.
  • ［B4］Theory possession: A state in which theory ceases to be a tool of the user, governs the user’s perception, self-understanding, and action, and tries to make reality conform to itself.

Chapter 6: Gravity’s Rainbow: Paranexia through Excess of Connections

Connection produces a system. By relating part to part, an overall picture comes into view, and the world appears as something comprehensible. Scholarship is the institutional form of this activity. It connects fragments, describes relations, and constructs systematic frameworks. A properly functioning system works as an apparatus for capturing reality, like a spider’s web that catches prey.
However, connection too acts in the opposite direction when it becomes excessive. The more connections increase, the more complex relations become, and once a certain critical point is crossed, the system becomes impossible to grasp. As this progresses further, the system becomes not an apparatus for capturing reality but itself a self-proliferating labyrinth. Building the web, moving through it, and the act of extending it themselves become ends in themselves, and ultimately the one who produced the system is trapped by their own system. This chapter describes this reversal phenomenon as Paranexia.

6.1 The Reversal of Connection: From System to Labyrinth

At the stage where connection is functioning normally, relations between elements help in grasping the overall picture. A map shows the territory, a systematic framework explains phenomena, and theory makes prediction possible. Connection is a tool of understanding. This normal function is established only at a stage where the increment in understanding obtained through connection surpasses the increment in the cost of grasping that connection.
When connection becomes excessive, this relation reverses. As elements increase, relations between elements increase, and relations between relations increase, the overall picture is not updated and only the cost of grasping it increases. What is occurring here is not ignorance. It is the paradoxical state in which explanations increase but the overall picture cannot be drawn. The increase in connections appears not as an increase in understanding but as the saturation of processing prior to understanding.
Next, connection takes on the character of paranoid associative delusion. Everything begins to appear related, relations call forth relations, and meanings produce meanings. Connection itself becomes an end in itself, and the system is transformed from an apparatus for capturing reality into an apparatus for proliferating connections.

6.2 Gravity’s Rainbow: Anatomy of a Broken System

Thomas Pynchon’s Gravity’s Rainbow embodies this process in literature. A vast number of elements are scattered throughout the work. The trajectories of V2 rockets, networks of secret organizations, experiments in Pavlovian conditioning, sexual anecdotes, Kabbalistic numerology, equations of thermodynamics, mythic fragments. These are connected, refer to one another, and exchange meanings.
In many mystery-solving and teleological novels, the system moves toward convergence. Clues move toward the solution of the mystery, characters move toward their destinies, fragments move toward the whole. But in Gravity’s Rainbow, every time a clue is presented, another clue intrudes with equal weight. Connections that should converge on a center continue to produce other connections. The reader is entangled in the web of relations stretched throughout the work and yet cannot grasp the entirety of that network.
The system of Gravity’s Rainbow is a web of derangement. The web exists. The threads have been stretched. But the function of catching prey has been lost, and only the extension of the web continues. The outlines of the conspiracy are intimated but not settled.
What is losing its function here is not the work but the explanatory system constructed within the work. By retaining the dysfunction of that system as form, the novel continues to operate at another level.

6.3 The Transformation of Means into Ends and the Self-Capture of the System

There are stages in the progression of Paranexia.
The first stage is the normal proliferation of connections. Elements increase, relations increase, and the system’s capacity to capture increases. The larger a spider’s web becomes, the wider the range it can capture.
The second stage is the crossing of the critical point in the cost of grasping. Connections increase too much, making the overall picture impossible to grasp. The web becomes too large, and it becomes impossible to grasp the whole of it as a single trapping apparatus.
The third stage is the transformation of means into ends. When the overall purpose becomes impossible to grasp, the only thing that can be measured is the local increase in connections, and connection itself produces a meaning reward. Proliferating the system takes priority over understanding reality. Extending the web takes priority over catching prey.
The fourth stage is the stage at which the system traps the one who produced it. The proliferated system constrains the very person who produced it. The scholar is exhausted by maintaining systematic frameworks, the bureaucrat is buried in procedures, and the system ceases to be a tool for the one who produced it and continues its internal proliferation.
This fourth stage is the endpoint of Paranexia. Without fully collapsing, the system repeatedly self-proliferates and reaches dysfunction.

6.4 Paranexia in Individuals and Groups

Paranexia does not occur only in collective systems. An isomorphic collapse also occurs in individual thought.
At the individual level, it appears as associative-delusional thinking. Every event seems to be connected to every other, relations call forth relations, and interpretations produce interpretations. Thought becomes not a tool for grasping reality but an apparatus for proliferating connections. Just as in Pynchon’s own works, and just as in the experience of Slothrop, the protagonist of Gravity’s Rainbow, Paranexia turns individual recognition into a labyrinth.
At the collective level, the specialization of scholarship localizes networks of connections, and new knowledge and institutions created to connect fields further increase the total amount of connections. In organizations as well, meetings, documents, and approval routes established to connect departments become new objects of coordination, and maintaining connections becomes work.

6.5 The Structure of Paranexia

Paranexia is a state in which, through the proliferation of connections, the outline of a system is lost, the system is transformed from an apparatus for capturing reality into a self-proliferating labyrinth, and ultimately traps the one who produced it.
Self-capture appears as a collapse in grasping, in which no one can grasp the overall picture of the system. Local processing may still function, but that locality makes the outline of the whole even harder to see.
What is distinctive about Paranexia is that the system does not disappear. Connections are maintained, local processing continues, and new relations are added. Nevertheless, the system becomes not a map for understanding the world but a maze that continues to postpone understanding.
Some spiders that spin orb webs eat their old webs and spin them anew. Human system-building has neither such a discipline of dismantling nor a natural upper limit. The endpoint of Paranexia is not the vanishing of the system. The exit does not disappear. Too many routes arise, making it impossible to determine which route leads to the exit.

［Notes］

  • ［B3］Pynchon’s works are sometimes read as “system novels,” but the system that arises here is not refined like a scholarly systematic framework. Because connections alone proliferate while lacking a center and boundaries, the system becomes a bundle of tangled threads and leaves the reader not a map but a maze. Gravity’s Rainbow is known as an “unreadable novel” not simply because it is difficult, but because the system does not properly close as a system.
  • ［B1］Paranexia: A compound of the Greek para (abnormality, deviation) and the Latin nexus (connection, entanglement). A state in which the outline of a system is lost through the excessive proliferation of connections, and the system turns into a self-proliferating labyrinth. The analogy with paranoid associative delusion is intentional and implies a structure in which the state of seeing everything as connected produces dysfunction in the system.
  • ［B3］Conspiracy theory is a typical social case of Paranexia. An interpretation in which every event is connected by a single hidden intention appears as the excessive proliferation of connections. Individual connections have meaning locally, but the system as a whole becomes an unverifiable labyrinth. Conspiracy theory is called a closed system, but structurally the reverse is the case. Because it continues to absorb new evidence as connections, it closes epistemically.
  • ［B3］The specialization of scholarship can be observed as an institutional form of Paranexia. The more fields increase and connections increase, the more the outline of scholarship as a whole is lost. Interdisciplinary research arose as a response to this problem, but there is a paradox in which interdisciplinarity itself produces new connections and further proliferates the entanglement. An attempt to repair the system deepens the system’s entanglement.
  • ［B3］Human beings become entangled in the networks of relations they themselves produced. This is not an individual failure but a structural consequence inherent in the act of connection. Building a system and being trapped by a system are the two sides of the same movement.
  • ［B4］Does a reader who understands Gravity’s Rainbow really exist? Honestly, I have to admit that I don’t understand Gravity’s Rainbow at all. Its characters number more than four hundred; its contents range from ballistics and rocket engineering to the international political situation, conspiracy theory, vulgar pornography, and occultism; and, what is more, these elements are entangled with one another. It is not merely that the work is enormous; it exceeds the amount of information a single human being can understand. Extract a part from the whole and that part, as far as it goes, can be understood, but understanding the work as a whole is impossible.
    Two questions surface here. One is the question, “Can one analyze, criticize, and evaluate a work one does not understand?” And the other is, “Might Pynchon himself not understand his own work?” I answer the former. One can. Understanding is the enemy of criticism. Every reader who claims to understand Gravity’s Rainbow is deranged. It is like a sick person who, in the same instant as going mad, opens their eyes wide, raises both hands, and bursts out screaming, “I have understood the truth of the world!” As evidence: the famous Gravity’s Rainbow fan site resembles a paranoia patient’s homepage. Everything is given a meaning, everything is connected to everything else, and yet the whole cannot be surveyed. It looks less like a guide to literature than like an archive in which paranoiac reading has organized itself. Next I answer the latter question without showing grounds. Pynchon does not understand Gravity’s Rainbow. Nor does he control it. In short, I am diagnosing both the author and the readers of Gravity’s Rainbow as paranoiacs. This diagnosis is not my invention. In 1974, the Pulitzer Prize fiction jury unanimously recommended the book, but the advisory board rejected it as “unreadable, turgid, obscene, and overwritten,” and that year’s prize was brought to a close with the winner’s place left vacant. The advisory board’s findings of fact are entirely correct. What was mistaken was only the sentencing. To be unreadable, turgid, obscene, and overwritten—that is not a flaw in this work of art.
    Art, as meant here, refers not to the completion of mastery but to that which arises in the tension between mastery and breakdown. A masterpiece and a botched work are not different in kind. Both are born of the same excess and share its nature indiscriminately. A work that intention and technique have mastered down to every corner remains craft, however exquisite it may be. The artist’s ambition moves from the controllable safe zone toward the uncontrollable danger zone. An uncontrollable rocket flies until its fuel runs out.
    That said, a work that has merely lost mastery is nothing more than a simple runaway, and it lacks tension just as the well-made work does. That which is collapsing and yet does not finish collapsing. That which, still carrying breakdown as movement, does not complete its fall. The way the rocket freezes above the roof of the theater, in the last immeasurable gap—falling, never reaching impact. Only in that suspension—or not even in suspension: only in a state of contradiction in which formation and disintegration, continuity and discreteness, hold simultaneously—does the domain of literature open. What is broken in Gravity’s Rainbow is the system stretched throughout the work. Slothrop’s erections anticipate the impact points of the V2s falling on London. Response precedes stimulus, and the system of causality is standing on its head. Pointsman’s Pavlovianism and Mexico’s Poisson distribution are brought in to explain this headstand, and both fail alike. And then in Part Four, the protagonist himself, who was supposed to be at the center of the system, begins “to thin, to scatter,” becomes unable to maintain the unit called a character, and the story proceeds to its end with the protagonist having exited. The web is not functioning in any normal sense. But the work, even as it expands the system and stages collapse, never entirely loses tension, and is just barely standing. That fat, bloated pig, which tries to swallow the entire world and to enlarge its own description until it becomes the world itself (borrowing Henry James’s words, one may also call it a “large, loose, baggy monster”), is doing a handstand on a single foreleg, and running through it is a contradiction that cannot be accounted for by the words chaos and order. At the opposite pole stands Flaubert. He dreamed of “a book about nothing,” of a book that would float in midair on the internal dynamics of style alone, and with Madame Bovary he formally achieved that reverie. Its prose is thoroughly mastered from beginning to end. Madame Bovary is the finest piece of craft humanity had reached at that point, and for that very reason it is not art. Pynchon’s Gravity’s Rainbow stands at the opposite pole from an anti-flab aesthetics like that of Flaubert’s Madame Bovary. But that does not mean there is no tension there. It gives form to an enormous world with a tension of another kind.

Chapter 7: Infinite Jest: Anakleisis through Excess of Recursion

Recursion is, in principle, an apparatus for deepening thought and action. Human beings can distance themselves from a simple chain of reactions and reach higher-order judgment by looking back on themselves, examining their own actions, and reflecting on experience. Recursion is a condition of self-correction and maturation.
However, recursion too brings about the opposite effect when it becomes excessive. Reflection becomes not something for guiding action but a closed circulation, and thought continues to repeat the same path. Self-consciousness, far from helping self-governance, draws resources into maintaining self-consciousness itself. Repetition is transformed not into proficiency but into constraint, introspection not into maturation but into inescapable hesitation. This chapter describes this state as Anakleisis.
Running but not advancing. In motion but not moving. Like a mouse that continues to run in an exercise wheel, the subject in Anakleisis operates actively while doing nothing but repeating the same path.

7.1 The Reversal of Recursion: From Reflection to the Closed Loop

At the stage where recursion is functioning normally, the subject looks back on itself and modifies its actions as a result. Thought returns to reality, and reflection makes the next choice more appropriate. Recursion is an open circuit, and the inner life forms part of a back-and-forth movement with the external world.
When recursion becomes excessive, this back-and-forth movement turns into a closed loop. As thought calls forth thought, reflection calls forth reflection, and self-consciousness amplifies self-consciousness, the subject can no longer return to reality. Thinking about a certain problem is transformed from a means of solving that problem into an apparatus for sustaining the act of thinking. Thought is occurring, but that excessive thought appears not as forward movement but as circulation.
As this progresses further, recursion becomes linked with dependence and compulsion. The subject repeats in order to escape pain, deepens the pain further through repetition, and repeats again in order to escape that pain. A closed circuit of circulation is formed, and action does not move toward a purpose but merely calls forth the next action.
Within this closed loop, continuing to think deepens the closed loop, while attempting to stop thinking makes “stopping” a new object of thought and returns to the closed loop. Either choice returns to the same constraint. And as will be seen below, this constraint is not a single circuit. It appears as heterogeneous circuits at the level of the subject (7.2), the level of mediation (7.3), the level of textual form (7.4), and the level of linguistic institutions (7.6). Anakleisis is the general name for states in which, within these various circuits, an exit comes to appear only in the form of repetition, ritual, or surrender.

7.2 Hal Incandenza: The Closed Loop of Intelligence

In David Foster Wallace’s Infinite Jest, the most precise depiction of Anakleisis is concentrated in the character Hal Incandenza.
Hal is a tennis genius who possesses exceptional intelligence and linguistic ability. He commands an enormous vocabulary, develops complex thoughts, and can analyze his own condition with great precision. Yet this intelligence and capacity for introspection function as an apparatus that severs Hal from the external world.
The college admissions interview placed at the beginning of the work is a condensation of this structure. Internally, Hal thinks clearly and prepares appropriate responses. Near the very beginning of the scene, the sentence he utters in inner speech, “I am in here,” is a declaration of the existence of an inner life that does not reach the outside. But to the ears of the interviewers, his utterance arrives as sounds that are not even language, as animal howls. Language is operating excessively within Hal, but that language has lost its output circuit toward the outside. Thought circulates within thought and does not reach the outside. The paradox in which the deeper introspection becomes, the more connection with the outside is lost, is made visible in this single scene.
However, this should not be read as a simple rupture between a rich inner life and impoverished output. What is damaged in Hal is not the inner life but the circuit of communication itself. When the desire to communicate passes through excessive self-consciousness, that self-consciousness burns out the organ of communication. The blockage is an event inside consciousness and, at the same time, is witnessed from outside as bodily transformation—howling, convulsions, animalization. He is not being misunderstood. The will to communicate is incarnated as the impossibility of communication.
What must be noted is the fact that, although this scene is placed at the beginning of the work, chronologically it is the final point of the entire story. Infinite Jest begins from its own ending. What the reader first witnesses is Hal already at the bottom of the closed loop, and the more than one thousand pages of main text retroactively fill in the process leading to this endpoint—while leaving one decisive point missing. The causal link that plunged Hal into this state is written nowhere in the main text. It has fallen into the gap between the final page and the first page. The meaning of this formal operation will be discussed in 7.4.
Hal’s problem is not a lack of intelligence. It is the excessive inward circulation of intelligence. He can analyze his own condition. As he analyzes it, that analysis produces the next analysis, and the circulation of analysis substitutes for action. Self-consciousness should make self-governance possible, but the excessive operation of self-consciousness constrains the self. Recursion becomes not a condition of maturation but a form of blockage.

7.3 “the Entertainment”: The Limit of the Closed Loop

The title Infinite Jest refers to a film left behind by Hal’s father, the filmmaker James O. Incandenza. Called “the Entertainment” or “the samizdat” in the work, this film strips anyone who watches it once of every motivation. The viewer loses interest in every action other than watching it again—eating, sleeping, conversation—continues watching the same images in front of the screen, becomes a wreck, and dies.
This is the literal materialization of Anakleisis. The circuit of consumption in which pleasure calls forth action and action calls forth pleasure normally keeps the subject alive because it is incomplete. Pleasure decays, boredom arises, and the subject is pushed outside the circuit. “the Entertainment” is an apparatus that perfects this circuit. Pleasure that does not decay, stimulation that produces no boredom. When the gain of the circuit becomes one or greater, the loop absorbs all of the subject’s resources and closes. This is also the general critical point of Anakleisis. If the gain of recursion is less than one, the circuit decays and returns to reality; when it becomes one or greater, it shifts into a closed loop in which operation reproduces the next operation. Perfect entertainment is the concept of entertainment carried out to its limit, and its completion coincides with the death of the consumer.
According to testimony within the work, the film imitates an infant’s vision through a special lens and places a maternal figure in a position looking down into a cradle. She bends down and repeats words of apology without limit. The substance of perfect pleasure is regression. At the bottom of the closed loop lies the position of an infant who need not respond and merely receives. Moreover, the content of the film is never presented directly in the work even once and is always given only through layers of hearsay and testimony. This blank belongs to the theoretical consistency of the apparatus. That which would draw the reader in if represented can exist within the work only as a blank in representation.
However, the core of this apparatus lies in the motive for its production. According to James O. Incandenza himself, speaking as a ghost in the work, the purpose for which he made the film was to draw his increasingly inward-closed son Hal back outside. To strike a son whom no words could reach with a work possessing an irresistible attraction and drag him out from inside his silence. In other words, “the Entertainment” was designed as a therapeutic apparatus. Something made to draw the subject out of a closed loop is completed as the most perfect closed loop of all. The attempt at recovery produces the worst structure of dependence—the central dynamics of Anakleisis are given narrative form here as the history of a single film.
It is not accurate, however, to read this apparatus as a general criticism of media. Lines of responsive mediation remain within the work. The voice of late-night radio, for example, is one-way yet does not constrain its listener and functions as a circuit of anonymous intimacy. Lethality belongs not to mediation but to a form of mediation that structurally blocks the possibility of response. “the Entertainment” requires no response whatsoever from the viewer and is altered by no response whatsoever. Perfect closure is another name for the complete absence of dialogue.
Furthermore, this film is pursued as a weapon of anti-American terrorism by a group of Quebec separatists in wheelchairs. It is a weapon that requires no coercion. The viewer chooses the loop voluntarily. A society whose organizing principle is the pursuit of pleasure structurally possesses no immune system against perfected pleasure.

7.4 The Closed Ring as Form: A Novel That Begins at the End and Sends the Reader Back to the Notes

The closed loop in Infinite Jest does not remain at the level of its characters. The form of the work itself is designed as a closed ring.
First is the temporal structure. As described above, the novel begins from the chronologically final scene. The more than one thousand pages of main text then end by cutting off, immediately before resolution, the central questions of the cause of Hal’s collapse and the whereabouts of “the Entertainment.” The ending does not exist inside the main text. It lies beyond the final page, which is to say, before the first page. To close the story, the reader must return to the beginning. Finishing the book is not completion but the beginning of rereading. The text keeps the reader on the ring and does not let the reader off.
Second is the apparatus of the notes. Infinite Jest has 388 endnotes, several of which carry footnotes of their own. The notes are isolated at the end of the volume, and the reader is forced to continue physically moving back and forth between main text and notes. Two bookmarks, the movement of a hand going back and forth. The act of reading itself is organized not as forward movement but as back-and-forth movement. The reader does not read Anakleisis. Through reading, the reader is made to enact it.
Moreover, the notes are not neutral additions of information. A considerable quantity of them is excess that does not contribute to the resolution of the story—a pile that includes refuse that ought to be screened out—and the reader is continually forced to distinguish the important from the useless. And when the reader returns from a note to the main text, the reader does not return to the same place from which they departed. Information that arrives belatedly retroactively rewrites the main text already read. The back-and-forth movement repeats the same path, yet with each lap rearranges the whole of understanding.
In this respect, Infinite Jest does not merely describe collapse as a theme; it is an apparatus that transcribes the form of collapse into the act of reading. In the terms of Jōkai Theory, this is a displacement from the level of description to the level of enactment, and is one of the most thorough implementations among the works treated in this paper.
There is, however, a designed asymmetry within this closed ring. The point of narrative convergence has not been deleted but relocated outside the main text—into the reader’s work of integration. Closure is not given; it is imposed. In contrast to “the Entertainment,” which kills its viewer through perfect closure that demands no labor whatsoever, Infinite Jest as a book is constructed so that it can close only through the reader’s active work. The gap is not a defect. It is an antidote, at the level of form, to the MacGuffin that this book carries within itself. It makes the reader run on the closed loop while, through that running, forming techniques of attentional selection and integration—the enactment of the closed loop and the training of resistance to the closed loop take place simultaneously on the same apparatus.

7.5 Inward Circulation Rather Than Outward Diffusion

In this respect, Infinite Jest stands in contrast to Gravity’s Rainbow. In Gravity’s Rainbow, connections proliferate farther and farther outward, and the system loses its outline and is transformed into an entanglement. It is a web of derangement produced by a spider: the threads are stretched, but it does not function as an apparatus of capture. What is there is Paranexia, the outward collapse of a system through the excessive proliferation of connections.
By contrast, in Infinite Jest, the problem lies not in outward diffusion but in inward circulation. Thought repeats the same path like a mouse in an exercise wheel, and action reinforces the closed loop. Rather than expanding outward until it becomes impossible to make sense of, it gradually comes to a stop while continuing self-referential repetition within itself. If the former reaches inability to be brought under control through proliferation and branching, the latter reaches inability to exit through repetition and recursion. In terms of the contrast between an explosive loop and an infinite loop, Paranexia is an explosive loop and Anakleisis is an infinite loop. Both continue to progress, but the former spreads outward while the latter closes inward.

7.6 Counter-Loop: Gately and Dead Language

There is another reason why Infinite Jest does not remain merely a description of a case of Anakleisis. The novel implements within itself, as a counter-apparatus, a condition for exiting the closed loop: Don Gately, the recovery facility in which he lives, and AA (Alcoholics Anonymous).
Gately, who is in the process of recovering from drug dependence, is placed at the opposite pole from Hal. Where Hal sinks through analysis, Gately rises by blocking analysis. The descriptions of AA in the work consistently repeat one principle. Do not ask why the program works. Just do it. Kneel, pray to a higher power you do not believe in, recite clichés. One day at a time. Keep it simple. The single term “Analysis-Paralysis,” which appears as part of AA vocabulary in the work, condenses the novel’s central diagnosis that examination through intelligence itself is what kills the dependent person.
Here lies a distinction decisive for the theory of Anakleisis. AA treatment does not eliminate repetition. On the contrary, it prescribes repetition. Daily meetings, the same clichés, the same rituals. What is removed is not repetition but the self-reference that has possessed repetition. A cliché is intentionally killed language. Language whose information content has reached zero offers analysis no purchase whatsoever. Because one cannot think about it, it can operate outside the closed loop of thought. The novel itself formulates this paradox—the vapider the cliché, the sharper the canines of the truth it covers. Meaning here arrives not through refinement but only through becoming banal. If malignant recursion is the fusion of repetition and self-reference—a structure in which reflection calls forth reflection—then treatment is surgery that cuts self-reference away from repetition.
The limit of this counter-apparatus is placed at a hospital bed near the end of the novel. Gately, shot while protecting a resident of the facility, remains intubated and without a voice while continuing to refuse narcotic painkillers offered by doctors as legitimate medical treatment—the very object of his former dependence. There is no heroic single blow that breaks the closed loop. What exists is an endurance stripped of self-reference: dividing the time that must be endured down to the granularity of single instants and continuing to repeat only the present, connected neither to regret over the past nor to projection into the future.
The condition for exiting Anakleisis is therefore not the stopping of repetition. To ask the dependent subject to stop repetition is equivalent to making another closed loop inside the closed loop—a closed loop of continuing to think about stopping. The condition of exit is repetition from which self-reference has been severed, that is, substitution with ritual. Infinite Jest implements this distinction over a thousand pages as the contrast between the two trajectories of Hal and Gately.

7.7 Anakleisis in Individuals and Groups

Anakleisis does not occur only within an individual’s inner life. At the individual level, it appears as compulsive repetition, dependent behavior, and excessive self-consciousness. By looking at itself too much, the subject moves away from action and turns the very process of attempting recovery into a new closed loop.
An isomorphic blockage also occurs at the collective and institutional level. Institutions that existed in order to describe and intervene in the external world begin to repeat only self-inspection, self-defense, self-annotation, and self-maintenance. Scholarship makes scholarship itself its subject rather than the world, while in organizations, meetings, evaluations, improvements, and procedures that had been means for achieving goals become ends in themselves.
What is specific to Anakleisis is not that elements proliferate outward, but that the same operation folds back onto itself. Reflection produces reflection on reflection, improvement produces improvement of improvement procedures, and efforts to get out of the closed loop add another closed loop.
When movement that should be directed outward continues to circle internally, the system operates actively while yielding nothing. Running but not advancing. The exercise wheel from the beginning of this chapter is installed not only inside the individual’s skull but also in institutional conference rooms.
As an exit for the individual, Infinite Jest presents not the stopping of repetition but the cutting of self-reference away from repetition. Ritual rather than analysis, endurance one day at a time rather than conquest of the future. But for collective closed loops, the work does not guarantee the same exit. Only the men kneeling on the floor of the facility and reciting clichés to a God they do not believe in faintly point in that direction.

［Notes］

  • ［B1］Anakleisis: A compound of the Greek ana (again, backward) and kleisis (closing; the noun form of the verb kleiein 〈to close〉). A state in which the excessive proliferation of recursion causes thought, action, and desire to form closed circuits of circulation from which the subject becomes unable to exit. Whereas Paranexia describes the outward collapse of a system, Anakleisis describes the inward blockage of a system. The two are contrasting types of collapse brought about by excess. Although the term was coined as a word distinct from the psychoanalytic term anaclisis (reliance, leaning-on), it is phonologically and semantically close to this chapter’s theme of infantile dependence.
  • ［B3］It is not accidental that addiction-recovery communities such as AA play an important role in this work. Recovery can be read as an attempt to exit Anakleisis. But the recovery Wallace depicts is not the elimination of circulation but the exchange of one type of circulation for another. While a self-referential attempt to stop repetition reinforces the closed loop, repetition from which self-reference has been severed can become a condition for exiting the closed loop. This distinction shows the essence of Anakleisis.
  • ［B2］The title Infinite Jest derives from “a fellow of infinite jest,” spoken by Hamlet in Act V, Scene 1 of Hamlet as he recalls the dead jester Yorick.
  • ［B3］The title “infinite jest” implies the structure of Anakleisis. A jest normally ends with a punch line. An infinite jest has no punch line. Laughter merely prepares the next laughter, and the circulation does not close.
  • ［B3］Anakleisis and benighted erudition describe different phenomena but readily operate in conjunction. The more knowledge proliferates, the more self-referential circulations of knowledge proliferate, and the more circulations of knowledge proliferate, the more difficult output toward the outside becomes. Hal’s closed loop of intelligence can be read as an individual form of this conjunction.
  • ［B3］Two faces and one spirit of the age—Gravity’s Rainbow and Infinite Jest can be read as two faces, mania and depression, of a single spirit called the present age. Mania and depression are not two different illnesses. One excess becomes mania when it leaks outward and depression when it encloses itself inward. The former proliferates outward as excess of connections and becomes a labyrinth that has lost its center (Paranexia). The latter revolves inward as excess of recursion and becomes a closed loop that has lost its exit (Anakleisis). They are the centrifugal and centripetal movements of the same excess.
    What must be noted is that neither work is a pure pole. Gravity’s Rainbow rises manically while falling, pulled by gravity, into entropy and death—the arc of mania bends directly into a depressive fall. Infinite Jest contains the depressive themes of addiction and paralysis within the manic proliferation of swelling endnotes—the content of depression floods in manic form. That each work carries its own opposite is not the breakdown of this reading but evidence for it. Under sufficient pressure, one pole turns into the other. This is why the two are two vectors of a single excess.
    The present age is a state in which this excess cannot remain still in one direction and continues to oscillate between outward proliferation and inward rotation. The two novels are two portraits of such movement. This is not proof that the present age is bipolar. It is a case in which the bipolarity of the present age has become readable in two works.
  • ［B3］Cynicism and sincerity—placing the previous chapter’s Gravity’s Rainbow beside this chapter’s Infinite Jest, one notices that the position of laughter is exactly opposite. Pynchon’s laughter points outward. Authority, the establishment, bureaucracy, “They”—the target of ridicule is always outside the self. Wallace’s laughter points inward. The laughing self itself becomes the object of surveillance, and the suspicion “Isn’t this laughter a defense?” follows laughter like a shadow.
    This turn has been described as an opposition: postmodern cynicism—the gesture of placing oneself in a position that sees through everything and preemptively avoiding the danger of becoming earnest—and a new sincerity that overcomes it. Wallace wrote this configuration in a form that decisively influenced how it would later be read: cynicism was no longer a weapon for exposing authority but a gesture of consumer culture; the next rebels would be those who spoke without fearing naivety; discard cynicism and choose sincerity.
    However, cynicism and sincerity are not two opposing attitudes. They are merely different directions of the same single function—criticism. When criticism points outward, it becomes cynicism. It exposes the deception of others and treats authority with contempt. When criticism points inward, it becomes sincerity. It exposes the deception of the self and continues to doubt itself. The extreme of the former is contempt for others, and the extreme of the latter is self-abasement. Both perform the same operation: exposure.
    The turn from Pynchon to Wallace therefore amounts not to throwing away the weapon but merely changing the direction of its muzzle. What Infinite Jest depicts is a record of what becomes of criticism directed inward. Sincerity that doubts the self goes on to doubt the doubting self, and then doubts that doubt again. Sincerity was not an exit from cynicism. It was the centripetal phase of the same disease.
    Just as mania and depression are the centrifugal and centripetal movements of one excess, cynicism and sincerity are the centrifugal and centripetal movements of one criticism. Two terms that appeared to stand in opposition fold into one.
  • ［B4］In the annotation to the previous chapter, I admitted that I don’t understand Gravity’s Rainbow at all. This time I will make an even worse admission. Since this is an outright crime in critical activity, it would be better to rename it a “confession.” But if a confession is to be proof of sincerity, then the greater the crime, the greater the evidence of sincerity it would make. I have not read Infinite Jest. I have not read a single page. The description in this chapter was written through dialogue with AI claiming to have read the work. Hal’s interview scene, the exercise wheel—I have confirmed none of it in the original text. The number of endnotes, though—that alone I know. 388.
    Here a question surfaces again. “Can one criticize a work one has not read?” In the previous chapter, to “Can one criticize a work one does not understand?” I answered instantly: “One can.” This time I cannot answer. I don’t know. If, hypothetically, this Chapter 7 were the best made of all the literary chapters, nothing would be more humiliating to me. The final judgment on this point I leave to the reader.
    I will attempt a defense. First, I am Japanese and can read only Japanese. The work exceeds a thousand pages. And no Japanese translation of Infinite Jest has been published. There is no way I could read it. Second, even if it had been translated into Japanese, I would probably be unable to read this highly intellectual work to any satisfactory degree. As for the first and the second, the cause lies not merely in my being unserious—even granting that this, too, could be counted among the causes—but simply in the intellectual capacity of the critic, myself.
    Third, as to how to talk about books you haven’t read, there is excellent prior research by Pierre Bayard. I have not read this either. Fourth—while thinking up a fourth defense, I notice that the defense has begun to call forth the next defense. I stop here. For the record, this is not irony in the classical sense (feigning ignorance). I really haven’t read Infinite Jest, and I can’t read it.
    As I write this, I am remembering a certain incident. It was in my student days, when I submitted a report for a literature course. The self of those days was an ardent literary youth. Though I had only four friends at the university, within that circle I was recognized as the one who knew literature. So when a report was assigned in the literature course, my spirits ran high. I thought it the perfect chance to bathe in the limelight. I scrawled down what I had felt and thought about the work, bundled it into 30 sheets of A4 paper, and submitted it. Days later, when the professor summoned me, I opened the door with my chest dancing with anticipation, convinced that I was going to be praised. Far from being praised, I was beaten to a pulp. At first, in a gentle tone (a tone with which one might comfort a child), he pointed out the problems in my report for me. At first I listened, feigning docility, but I had my own convictions where literature was concerned, and I could not keep myself from arguing back at him. Taking the form of questions at first, then turning to counterargument, I tried to argue him down. This cheeky attitude of mine ended up turning him into popcorn bursting open. Knowing no literary theory whatsoever, the self of those days was showered with them, completely undefended, over its whole body. That one I had not read, that one either I had not read, and this one I had failed to read—in short, that I was “self-satisfied and under-read” was pointed out to me in every variety of phrasing, and I suffered severe burns over my whole body. I have no intention, at this late date, of denouncing his words and conduct as academic harassment. Looking back even now, I feel it was good that the self of those days was beaten to a pulp then. In fact, I went on from there to study literary theory and became able, at a minimum, to handle it myself. The New Criticism, psychoanalytic criticism, postcolonial criticism, feminist criticism, structuralism, poststructuralism, etc. By the time I graduated from university I had also written five papers. But no one cites them, and they go unread. Even now, more than ten years after graduating, I still read criticism from time to time, but I cannot muster interest in the latest literary theory. From somewhere past thirty my interest faded by degrees, and lately I read almost nothing but writing on literature from before the New Criticism—that is, from before literature was theorized, institutionalized, and pedagogized. Benjamin, Baudelaire, and the like. I do not think their writing on literature holds good, unchanged, in the present day. Compared with criticism grounded in contemporary literary theory, their criticism is neither meticulous nor rigorous. There are leaps of logic, and in the scholarly sense it is not even accurate. But frankly, I am drawn to their words. Just as literature is not a matter of writing correctly, literature is not a matter of reading correctly. Correct, correct, this again, this again, correct, and yet this again… Theory brings repetition. Tedious. As for the problems inescapably bound up with literary theory, Part II of this paper enacts them, so there is no particular need to write them up here.
    Criticism by its nature has a side that criticizes institutions, and that function has been inherited by contemporary literary theory as well. But since the New Criticism, the aspect in which these theories have themselves become institutionalized cannot be ignored either. What the institution secures there is seriousness—scholarly earnestness—concretely, political correctness, logical correctness, and procedural correctness. These have significance, of course. But that such seriousness has significance does not mean that unseriousness has none. Nor does being serious necessarily surpass being unserious.

Chapter 8: The Mega-Novel: Pantrophia and the Post-less Era

8.1 The Inadequacy of the Term “Postmodern” and the Post-less Era

Contemporary literature has often been called postmodern. However, this term is not sufficient, at least for describing the mega-novel. This is because the present age is not an age in which a new stage beyond modernism has arrived, but an age in which, while aware of the limits of modernism, neither something to replace it nor something to follow it can be found, and it cannot be brought to an end. Exhaustion is acknowledged. The impasse, too, is registered. Nevertheless, no new principle appears that breaks through it. In this sense, what characterizes the present age is less that it is post than that the post does not come, that is, that it is post-less. This situation may also be called necromodern, in the sense that styles already dead continue to operate. Modernism, though it should have ended, does not fully end and continues to repeat its gestures and forms. But that repetition does not move toward the future. What is here is not revolution but prolongation. What post-less describes is the absence of an endpoint, an exit, and a successor. It is not closed in. It continues without being able to close.

8.2 Pantrophia

The failure of a new principle to arrive does not necessarily mean that form stops. Form tries to persist by internally proliferating, expanding, and densifying existing formal resources. Because it cannot go outside, it increases within. If a new style cannot be acquired, the only option is to take an already existing style to an extreme. What arises in this way is Pantrophia. However, hypertrophy cannot be said to be a symptom found only in the present age. A book such as Musil’s The Man Without Qualities, possessing an internal mechanism that might be described as generation principle > convergence principle, can reach its end only extrinsically. He did not lack time or skill. Rather, the more of them he had, the more additions were repeatedly made. With each revision, it did not move toward completion but expanded. There was more than enough, and yet only the end was absent. Therefore, it had no choice but to swell. Hypertrophy without an exit is not itself peculiar to the present age and can arise whenever particular conditions—more than enough, and yet no end—are present. Here, however, its characteristics will be described with contemporary Pantrophia in mind. Pantrophia in the post-less era is a state of expansion without innovation. Length, amount of information, number of characters, number of episodes, number of connections, number of annotations, density of style, and layers of narration increase, raising the density of the work as a whole. Instead of a new principle being born, accumulation and complexification proceed within existing principles. Pantrophia is not merely a quantitative increase. It is a state in which increase becomes a means for the survival of form, and then increase itself becomes an end in itself.

8.3 The Contemporary Mega-Novel Is Pantrophia in Literature

The mega-novel, a novel hypertrophied to its limit, constitutes an important extreme point within contemporary literature. But what deserves attention here is not in itself the characteristics of the mega-novel—that the work is enormous, that it is dense, that it is excessively connected, that it is encroached upon by annotations. The focus lies on the origin of the hypertrophy and on the desires, ambitions, and strategies imagined behind it. It is, so to speak, similar to describing under what constraints the style of the skyscraper arose, describing the origin of that arising. If construction is to continue within a limited site, the building rises upward instead of spreading sideways, and its interior becomes denser. High-rise development and densification can be redescribed not as the result of free expansion but as manifestations of spatial constraint. The same is true of the mega-novel. When a new literary horizon has not sufficiently opened, the novel is extended, multilayered, and hypertrophied within existing formal space. Enormousness as a sign of the post-less.

8.4 Aesthetic Shift: Pantrophia and Pantrophism

However, it is insufficient to understand being enormous, and an enormous style, merely as Pantrophian symptoms. This is because that hypertrophy can be regarded at once as a consequence of the absence of a successor and as an attitude, or a style, that attempts to encompass the complexity of the world nonreductively. The world is enormous, complex, polyphonic, fragmented, and yet these are connected with one another. Good and evil, high and low, center and periphery, scholarship and scandal, the sublime and the vulgar, institution and deviation exist there simultaneously. The world has another aspect that is difficult to grasp through simple and pure form. A style of description is not merely a method of expression but also a method of recognition. A mathematician produces new notation because there is a world that cannot be grasped except through that notation, and in this sense, the search for style means expansion in both expression and recognition. It is certain that enormous, dense, multilayered literary form gives shape to aspects of the world that cannot be sufficiently grasped by literary form that is not so. Pantrophia is at once a pathology and what might be called a clumsy response that tries straightforwardly to apprehend reality just as it is. Nor is extraordinary art necessarily a success. An artist has an ideal, but that ideal does not simply arrive at ideal completion; while struggling with reality, the ideal is distorted, bent, and wounded. There is creativity, necessity, persuasive force, and vitality that dwell in art produced from nonideal circumstances such as illness and disability or technical and economic constraints. Whether in the works of Picasso or those of Dostoevsky, traces of struggle with difficulties that resist subjugation are inscribed there. Works produced in this way cannot be the result of pure victory and at times contain both success and failure. Therefore, here the distinction adopted is not between success and failure but between symptom and attitude. Pantrophia, by itself, is a symptom and contains no will. But there can be an attitude that turns this symptom into affirmation and actively tries to take up hypertrophy as a style. This is called Pantrophism. This shift deserves particular attention. For example, if an obese child who simply cannot succeed at dieting, in one turn, makes the fullest use of that enormous body and becomes a sumo wrestler, then the child is no longer a “mere” obese child. At that point obesity remains obesity; that is, attitude does not improve the symptom, but once the child ties on a mawashi and steps onto the dohyo, the child is no longer merely a fatso but a rikishi. Even when faced with an unwanted situation or circumstance, even if it is an illness or disability, when the subject actively turns it into affirmation with a “Bring it on!” and steps onto the dohyo, the fatso becomes a rikishi, and Pantrophia becomes Pantrophism. The shift meant here is not to negate it as what it is, but, even if it is negated, to encompass even that and try to turn it into power, to turn it into a style, a strategy, a creed, an aesthetics. Not only the lazy fatso but also the diligent artist, architect, and entrepreneur are tossed about by a fate beyond their control. At such a time, rather than closing the mouth, they open it wide and try to swallow fate. This too is a characteristic of Pantrophism. Pantrophism does not adopt as its central principle the logic of exclusion according to which what is bad is stripped away and only what is good is left. True/false, beautiful/ugly, good/evil, sacred/profane, main line/digression, useful/useless—it absorbs, without distinction, as much as possible and turns everything into its own flesh and blood. This omnivorousness, this logic of inclusion, is the essence of Pantrophism. If a giant has any distinction-making, it is whether something can be eaten; if it can be eaten, it eats, and eating is neither affirmation nor negation. Once something falls into the stomach, is digested, and becomes flesh and blood, even distinction-making between friend/enemy and self/other becomes difficult, and the object ceases to be an object. It chooses to swallow it, the world, or to be swallowed by it. This is appetite, it is both domination and servitude, it can also be a contemporary strategy, it is an ambition guaranteed to break down—“to become the whole while being a part”—and it is a principle of love.

［Notes］

  • ［B1］Pantrophia: A compound of the Greek pan (all, whole) and trophē (growth, nourishment). A state of expanding without innovation while taking everything in through quantitative excess. It differs in character from the other concepts of excess in this paper in that it possesses the ambiguity of being readable both as a pathology and as a style.
  • ［B1］Pantrophism: An attitude that reverses the symptom of Pantrophia into affirmation, takes up hypertrophy as a style, and pushes it to its limit. Like accelerationism, Pantrophism is not a treatment for the disease. Whereas Pantrophia is an object of diagnosis, Pantrophism is an object of evaluation.
  • ［B2］The relationship between Pantrophia and the concepts of Chapters 6 and 7 can be organized as follows. Paranexia is the self-collapse of a system through excess of connections, while Anakleisis is the fixation of the closed loop through excess of recursion. Pantrophia functions as a condition that facilitates their occurrence. When form hypertrophies Pantrophically in the post-less era, Paranexia and Anakleisis become more likely to arise within it.
  • ［B2］Post-less lies close to “the slow cancellation of the future,” which Mark Fisher discussed by taking up Berardi’s phrase. Fisher argued that behind the superficial “newness” and movement of twenty-first-century culture lie anachronism and inertia, the repetition of past styles, and formal nostalgia. The state described in this paper, in which styles that should have died fail fully to end and prolong themselves while repeating their gestures, overlaps with this diagnosis. The level of explanation, however, differs. Whereas Fisher’s description moves toward cultural temporality, affect, and the haunting of lost futures, the post-less of this paper takes the operation of formal resources as its problem. When no new principle arrives, existing forms do not stop but persist through Pantrophian internal proliferation. Haunting here is formalized not as mere mood but as the expansion of length, amount of information, episodes, style, and connections. From the place where Fisher heard the voices of ghosts, this paper hears the hammering of extensions. The expansion of the post-less, moreover, resembles accelerationism but is not the same thing. Accelerationism was an eschatology that left faith in an exit within expansion, in that if a process were accelerated to its limit, something else—singularity, post-capitalism, completed capital—would arrive beyond it. But what arrived was not an exit but only scale, and under a regime in which everyone accelerates, acceleration is no longer a position but a condition. The guarantor of the exit continued to retreat from history to capital, and from capital to thermodynamics. Accelerationism was the last eschatology to mistake Pantrophia for a plan. A symptom, however, can turn into an -ism. An attitude that turns the absence of a counter-principle into affirmation and pushes hypertrophy itself to its limit—the Pantrophism of this chapter—is possible, and the wreckage of accelerationism can be read as its first form.
  • ［B1］Post-less: A state in which, despite awareness of the limits of modernism, a new principle or style to replace it has not sufficiently been established. What matters is not “that the next stage has arrived” but “that the next stage does not arrive.” Post-less is not stagnation but a historical state characterized by the absence of an exit.
  • ［B3］Excess in the post-less era is not limited to literature. In scholarship, institutions, legal systems, and technology alike, isomorphic Pantrophia readily arises at junctures where principled innovation is difficult. Complexification, accumulation, and densification proceed within existing frameworks, and quantitative excess functions as a substitute for innovation. This expansion does not arise from constraint alone. Just as high-rise development requires construction technology, Pantrophia arises where technological conditions that make quantitative expansion possible overlap with the impossibility of principled innovation. Generative AI is the extreme of those technological conditions, and the information explosion discussed in Chapter 11 can also be placed in this context.
  • ［B3］Maturity and ennui: Being enormous is not an invention of the post-less. As The Man Without Qualities shows, works also became enormous at the height of modernism. As stated in the main text, hypertrophy without an exit can arise regardless of period when the conditions “more than enough, and yet no end” are present. The symptom is diachronic.
    Post-less is not a periodization in the general sense. This paper does not mean to say that “the age of modernism ended and the post-less era arrived.” Rather, what is being said here is that “modernism did not end and continues without being completed, remaining incomplete, in a state in which revolution can no longer be expected.” In this sense, hypertrophy without an exit has existed since the age of modernism, and contemporary enormous styles can be situated on its extension.
    In the age of modernism, it was possible to believe in experimentation with form, the avant-garde, and innovation, but in the present age it has become difficult to believe innocently even in these. It is difficult not only to believe in the old, but also to believe in what is here now or in what ought to come next. What is here is the mood that “everything has already been done, perhaps there is nothing truly innovative, and even if innovation were possible, perhaps the landscape beyond it would not change very much.” Of course, this too is diachronic. Such a mood also existed in the age of modernism. At that time, however, it was still possible to envision an untouched outside. The mood of the post-less is the ennui that comes from being unable genuinely to believe in this reverie.
    Against this dominant contemporary mood, this sense of ennui, what can now be regarded as radical is no longer the revolutionary attitude of negating the present state and renewing it, but excess, acceleration, and expansion. If little can be expected from qualitative transformation, there is no choice but to expect a transformation from quantity into quality; put further, what is being envisioned at this point is emergence, or Jōkai.
  • ［B1］Necromodern: Because the post does not appear, the modern cannot die. Modernism, while exhausted by its own principles, cannot stop its forms and gestures. Democracy and capitalism too continue to repeat self-renewal and expand while undergoing dysfunction without reaching an end. Post, counter, and alternative alike are nothing more than movements not outside the system but inside it. Even declarations of death are encompassed by the narrative. Since modernity, deaths and endings have been shouted and whispered—the death of God, the death of man, the death of the author, the end of history, the end of philosophy, the end of art—but none has been able to die satisfactorily or to end. Dying but unable to die. Ending but unable to end. Being unable to end without an endpoint in sight, being unable to die while rotting, failing to place the period: these are characteristics of the post-less. It is not simply that they do not die. Even corpses are dug up and incompletely revived. Religion, tribe, monarchy, Nazism, Marxism… styles that should have died are dug up again and again, patched together, and brought back to life like monsters.
  • ［B3］Endless shōnen manga: Post-less is like a shōnen manga that cannot end. When a story is prolonged beyond the point where it should originally have ended, stronger enemies, larger settings, deeper pasts, and greater crises are added one after another. Each time, what should once have been the final juncture is demoted to an intermediate point. Moreover, in the process of prolongation, even enemies, settings, and origins that should have died are dug up, grafted on, and reused. A story that cannot end survives not through completion but through inflation. Post-less likewise prolongs itself not through ending but through expansion and reuse.
  • ［B3］The myth of creation: Modernity is negation, doubt, and overthrow through revolution. Behind it was a spirit seeking freedom. Rebellion through intelligence and reason against rulers killed God, killed the king, and killed the father. The sheep that killed the shepherd broke through the fence, expanded their territory without limit, and tried to govern themselves by their own power. It is a history of rule and invasion. The sheep fought, killed one another, and at last began to eat one another. The next sheep reflected on this tragedy. This time they began to doubt their own intelligence and reason. Endless self-reference and self-disintegration—from vertical struggle to horizontal struggle, from horizontal struggle to internal struggle. What matters is that modernism was not merely a reaction against institutions, authority, and tradition. They—the men and the women—rebuilt by themselves what they themselves had destroyed. We did not merely kill; we cut up the corpses and pieced them together, trying to resurrect a new God, king, and father. The above is a textbook historical view of modernity and modernism. But the myth described here is not something that can be seen only in one particular history. Whether they kill, die, or rise again, the sheep remain inside this circle, and they are often aware of it.
  • ［B4］While writing this paper, I have noticed something. Generative AI has a field at which it is better than human beings. That field is none other than institutional analysis, criticism, and description. What is institutional analysis, criticism, and description? That which has been theorized, methodized, and pedagogized. Criticism before the New Criticism, for example, depended on the critic’s intuition, sensation, and aesthetic experience—in short, on the bodily feeling of “i don’t know why but i felt it this way,” comically vague and, for that very reason, permitting no cheating. Of course, the critics of the past were also cunning and ignorant, swollen with vanity, trying to talk about themselves rather than the work, and repeating stock phrases all the while. I have no intention of uncritically extolling the atmospheric age when I was not yet even a sperm or an egg—a past that can be spoken of beautifully precisely because I cannot speak of it from experience. I do not mean to say that “school education has leveled individuality flat and a rich culture of reading has been lost.” That would itself be nothing more than stock discourse. Or perhaps, without noticing it, that is what I want to say. Knowing that it is stock nostalgia. There is no way to make sure of that without moving my pen further on. It is true that literature was taught before the New Criticism as well, but the education the New Criticism brought was not the mere transmission of knowledge about literature. The invention lay precisely in making the reading of literature a reproducible operation. One person’s reading could be reproduced by another using the same concepts, and the operation could be graded, compared, and corrected. The scholarly attitude that regarded “i don’t know why but i felt it this way” as an unscholarly and childish reading, placed it at the periphery of the institution, and pushed objective, lucid analysis, criticism, and description to the limit became dominant. The theorization, methodization, and pedagogization that the New Criticism brought to literature were of this kind, and this is institutionalized analysis, criticism, and description. What generative AI is good at is this kind of reading. Where theorized criticism is concerned, it already surpasses the majority of humans. The same goes for papers. It is very good at formal description and at evaluation based on its criteria. Institutionalized norms, moreover, are internalized by people. To give a concrete example, even the critical viewpoint that ［tightness, the absence of the author, organic unity, a structure without waste—the novel with an internal necessity in which every detail serves the whole is the 〈good novel〉, and Madame Bovary is the point of arrival of the modern novel］ can, without exaggeration, be traced in its institutional origin—the “institutional” origin of this aesthetic—to the New Criticism. The other wellspring readily named is James–Lubbock. Here I hear a voice crying “Objection!” Overruled. Kindly refrain from bringing up pre-nineteenth-century theories of drama, let alone Aristotle. Nor is there any need to touch here on the lineage of screenwriting manuals that flourished in the twentieth century. What would I be multiplying annotations for?
    In any case, it can at least be said that the New Criticism and the discourse around it did not merely present methods for analyzing literary works; they institutionally furnished literature with a certain anti-flab norm. Anti-flabism sounds abusive, but no better word comes to mind. I used the word in the annotation to Chapter 6 as well, and I have no intention, at this late date, of retracting it. What, roughly, is “institutional”? You may think it overblown. This is not a government office, after all—as if there were institutions for how literature is to be read and written. But there are. More than merely are. We read inside that institution, write inside that institution, praise and disparage inside that institution, and believe all the while that this is our own sensibility. Or rather, that would still be the better case: it has entered into us as doctrine. The theory of the New Criticism is dead. But that does not mean it has disappeared. Its critical viewpoint changes shape and, as a ghost, takes possession of our aesthetics, whispering to us with its name forgotten. Trim it down, trim it down. Even now, in fact, I am driven by an inexplicable sense of obligation, as if I had to trim the flab from this text. It comes from my left ear; from my right ear comes “Describe it rigorously.” And indeed, if I am to describe it more rigorously, loading all of these causes onto the New Criticism seems not only forced but unjust. If, for example, the women of a particular region in a particular period were considered more beautiful the thinner they were, and they really did suffer under that spell, it would in most cases be impossible to fix on any one point where the spell originates. Nor is the distinction between spell and aesthetics a certain one. That is what style is. Even the movement that tries to deviate from style is like that. Come to think of it, what lies at the base of generative AI’s evaluation of works is also the New Critical, or writing-workshop, viewpoint. Don’t let the author speak; show it in scenes rather than explanation; cut redundant description…—it corrects your manuscript straight out of the textbook. And if you instruct it to “criticize this from the viewpoint of the New Criticism,” it produces full-scale criticism. It can also handle theoretical criticism other than the New Criticism at a considerable level. Given the concepts, operations, and evaluative axes of a literary theory, reading a work according to that theory is something AI, even now, not infrequently does better than humans. In this paper, Chapter 7 is the working example. In other words, what is happening here is the mechanization of criticism. What has been theorized is methodized, and what has been methodized is mechanized. Where there are grading criteria there can be computation, and what can be taught can be imitated. It is easy to foresee a future in which criticism—theorized, methodized, pedagogized, institutionalized—is next mechanized. But this argument is no more than a leaf. Draw the leaf in, take hold of the stem, and try to drag it out, and at once the deep-rooted question comes into view: “Is there anything that will not be mechanized?” What, for example, cannot be substituted for by a machine? What has caught hold of me here is that which has not yet been born. What may happen next—the foresight of it—runs ahead of me and begins to curse me from the front. It has not yet arrived in complete form, but the premonition of its arrival keeps jabbing at us. This is neither Hiruko nor an ubume; it is the curse of the unbirthed, and the unbirthed’s gaze gazes at us from the future, suspecting that criticism based on literary theory may have been a half-mechanical operation from the start. And that eye pursues us from in front. The unbirthed is always ahead of us, keeping us under surveillance, with pupils like cameras, from a society where mechanization has advanced. What is an operation that is not mechanical? Say that AI cannot produce theory, and that too is doubtful. It is easy to dismiss AI’s products as mere recombinations of existing discourse, but human theory, too, is made of combinations of existing concepts, observations, metaphors, borrowings, and accidents. If recombination is the ground on which AI’s theories alone are called fakes, a considerable part of the history of human thought gets thrown out with them. Nausea rises; holding my mouth, I consider where this nausea comes from. The yellow lighting sways, losing its shape. The intellectual excitement and intoxication that progress brings, and an unstable sensation as if the ground were swaying underfoot; I try to grasp something certain to keep my balance, and the moment I grasp it I notice that its screws, too, are half pulled out, and I feel sicker still.
    “Humans have experience, have bodies, have souls,” I argued to S. Thinking back on it, this too seems a commonplace cliché, but being a cliché does not make it wrong. The discussion of Jōkai Theory, too, was mostly me talking; it was nothing that could be called a discussion. We were drinking at an izakaya in Ikebukuro, a few minutes’ walk from that Yodobashi Camera where the human-feces incident occurred. She put her hands on her hips and said she wanted to fix her slouch, and pointed out that my hair and beard had grown too long. I should never have drunk at all. With the momentum of the alcohol, there were things I said too much of. She was good at drawing people out, and I wanted someone to drag things out of me, so, rarely for me, I spat out everything, weak places included. By then, as I recall, I had already drunk three mugs of beer. She put both hands to her ears and laughed, showing her white teeth. From her thin lips came, in a small puff, the words “You’re still thinking about those things, are you?”—and though there should have been no poison mixed into them, for some reason I did not understand—of course, now I do understand—that offhand remark of hers opened my mouth wider still. I brought up Baudelaire and Benjamin, said “They never relied on anything like theory,” and went on about them for a while. From here on, I barely remember what I talked about. There is one image, and that image alone remains vivid in my memory and will not leave my head. In the wound on S’s pale breast, countless small crabs were teeming, and with their pincers they jabbed at her skin, or at innards that still seemed to be moving, cut pieces away, and carried them to their mouths. This came to me while I was relieving myself at a urinal. There was a strange nausea, and alcohol was not its only cause. The sounds—the voices above all—were deafening, and there was a sensation of those voices echoing and amplifying inside my head. This is an important part, so let me stress it: it is not that the nausea came after the image; without question, the image welled up after the nausea. S put money on the table, my share included, and dropped some coins as she did; I picked them up, and when I raised my face, under the yellow lighting, she had her black bag on her shoulder and was on the point of leaving. She was crying. I grabbed her arm and held her back. Losing what I might call almost my only friend was terrifying—the footing was dropping out from under me—and I was gripping S’s arm hard. It was after that. After that, thinking to take a deep breath, I headed for the toilet and saw the image in question. What was shameful was that I had certainly hurt S, yet could not remember what I myself had said. And even at that moment I was preoccupied with my own contradiction—that is, with the fact that I was denying theory while trying to construct a theory. When I returned to my seat she was not crying. To be exact, she was crying, but crying in a way that was not the tears I had seen a moment before. S had recovered her composure; she was not crying in the true sense, but seemed to be pretending to cry. It was at this moment—or perhaps it was not this moment, but at any rate somewhere in that day—that S wedged into a gap in the circuits of my thought a truly short phrase, and yet one I would never forget.

Part III: Operational Extension

Chapter 9: Slash-Maintenance Cost and Operational Bankruptcy — Verification Bankruptcy, Institutional Bankruptcy, Slash Melting

9.0 From Literary Case Histories to the Operation of Communities

Chapters 3 through 8 described the processes by which excess of language, rules, voices, connections, recursion, and form produces dysfunction in each locality. Literary works were not only materials for exemplifying Jōkai, but also case histories for observing junctures at which an increase in the aggregate produces dysfunction in an existing order.
From here, the unit of the question changes. What has been asked thus far is how much excess a single work or form can bear. What will be asked from here on is to what extent a community can continue to make distinctions among proliferating information, claims, evidence, rules, interpretations, and exceptions.
When the processing that a single work demands of its reader is demanded by society as a whole of its members and institutions, excess becomes not an aesthetic difficulty but an operational debt. From Chapter 9 onward, the mechanism of Jōkai is shifted from inside the work to the operation of communities, and the cost of maintaining the slash and the juncture at which that cost can no longer be paid are described.

9.1 Slash-Maintenance Cost and Critical Threshold A*

By slash, this paper means boundaries and distinctions such as true/false, fiction/reality, good/evil, human/machine, scholarship/art, and legal/illegal. Such slashes are often treated as things that exist naturally. But for a slash to operate effectively within a community, work such as determination, cross-checking, verification, authentication, enforcement, exception handling, and correction is required. A slash does not work merely by existing as an idea. It is repeatedly maintained through operation.
Let A(t) be the aggregate that the community must process. The elements here are the totality of things that become objects of slash maintenance, including information, claims, evidence, rules, interpretations, exceptions, documents, images, records, and participants. Following the definition in Chapter 1, A(t) includes not only individual elements but also relations among elements, competition, constraints, density of connections, and intensity of effects. As elements increase, the community must identify, compare, connect, rank, and, in some cases, exclude them.
The sum total of effort, attention, and human and technical resources that a community pays in order to maintain the slash is called slash-maintenance cost. Verification is required to distinguish true from false, institutional operation is required to distinguish legal from illegal, and authentication and identification are required to distinguish human from machine. The slash is not free. Moreover, as elements increase, comparison, cross-checking, relation-making, and exception handling among elements also increase, so maintenance cost can expand beyond simple proportionality.
Let A* be the upper limit of aggregate A(t) within which a community can stably maintain the slash. Within the range A(t) < A*, slash maintenance is possible even with friction and delay. But when A(t) > A*, the operational load exceeds the community’s tolerance, and slash maintenance becomes unstable. Excess is not simply a large quantity. It is the juncture at which the increase of the aggregate becomes too great relative to the community’s capacity to process it.

9.2 Operational Bankruptcy

What this paper calls communal intelligence is the totality of the community’s capacity to make distinctions in the world by drawing slashes and maintaining them. It is not the simple sum of individual intelligence, but operates through procedures such as verification, response, correction, and consensus formation.
A(t) > A* does not immediately mean slash melting. It is the critical condition under which slash maintenance becomes unstable and communal intelligence moves toward operational bankruptcy.
Operational bankruptcy is not a state in which the ideals or definitions of slashes have disappeared. Distinctions such as true/false, right/wrong, human/machine, and legal/illegal remain nominally. But the practical work of actually determining, verifying, enforcing, correcting, and sharing them no longer runs. Slashes continue to be spoken of, but at the communal level they do not operate sufficiently.

9.3 Verification Bankruptcy — Cgen < Cver

For a community to distinguish the truth or falsity of claims, it must invest time, attention, knowledge, institutions, and human resources in verification. But when claims, information, evidence, images, interpretations, and dissenting views proliferate excessively, it becomes difficult to verify them one by one. What occurs here is not the disappearance of truth but the breakdown of the verification process for arriving at truth.
This asymmetry can be expressed as follows.

Cgen < Cver

Cgen is the cost of generation, and Cver is the cost of verification. If the cost required for verification is higher than the cost required for generation, generation can get ahead of verification. However, this asymmetry alone does not cause bankruptcy. For individual claims, it has long been the norm that making something is cheaper than checking it. Verification bankruptcy appears as a juncture of its own when this asymmetry overlaps with an increase in A(t), and the total amount of verification cost required per unit time chronically exceeds the resources the community can invest. Cgen < Cver functions as an amplifier that converts an increase in A(t) into bankruptcy.
At this point, unverified things circulate provisionally, provisional things become established facts, and corrections are delayed. Verification bankruptcy is not a state in which falsehood has triumphed. It is a state in which the community, while still trying to distinguish true from false, can no longer make its verification processing keep pace with the volume being generated.

9.4 Institutional Bankruptcy — Cinst > Cinstmax

The same thing occurs with institutions. Institutional rules do not function merely by being written down. They operate only when accompanied by practical work such as interpretation, application, enforcement, reexamination, exception handling, and allocation of responsibility. But as objects, exceptions, domains of supervision, and involved actors increase, institutions take on complexity that can no longer be maintained.
This condition can be expressed as follows.

Cinst > Cinstmax

Cinst is institutional maintenance cost, and Cinstmax is the upper limit of resources that the community can commit to institutional operation. When Cinst exceeds Cinstmax, the community falls into institutional bankruptcy.
Guidelines, articles, regulations, and procedures remain. But the capacity to apply them consistently is lost. Even if rules are increased, operation cannot keep pace, and greater precision brings not greater robustness but greater fragility. Institutional bankruptcy is not the disappearance of rules but the loss of the capacity to make rules operate communally.

9.5 Nominal Slash and Effective Slash

An important characteristic of operational bankruptcy is that the slash does not immediately disappear. Truth and falsity continue to be spoken of, rules still exist, and differences between humans and machines are still asserted. But as the capacity to repeatedly determine, share, and enforce the slash declines, the discrepancy between the nominal and the effective widens.
Because the slash continues to exist formally, the community makes its own collapse difficult to see. But a slash being written down and a slash working are not the same thing.
Moreover, a hollowed-out slash does not stop charging maintenance costs merely because it has lost the function of distinction-making. Procedures of inspection, audit, compliance, and reporting continue to run, continuing to absorb attention and resources. Even as function approaches zero, the cost remains. After operational bankruptcy, slash-maintenance cost turns from an investment in distinction-making into pure loss. A slash that does not work costs more than the absence of a slash.

9.6 Phase Transition and Slash Melting

Operational bankruptcy does not proceed only as gradual deterioration. The phase transition described in one stage by the general model of Chapter 1 is decomposed into two stages at the operational level. When A(t) exceeds A*, the system undergoes a phase transition from maintenance of function to dysfunction. But the dysfunction here is not stoppage but hyperfunction marked by delay, provisionalization, and idle spinning (Chapter 2). For a time, the community can still maintain the slash on balance, even while carrying friction. But when accumulated bankruptcy reaches a second critical point—the reversal point—the direction of operation reverses. In what follows, critical point refers to this reversal point.
What occurs here is not a sudden fall into complete disorder. Verification, enforcement, correction, and judgment continue. Nevertheless, they are no longer steered toward maintaining the slash. The community continues to work, but the direction of its operation reverses from maintaining the slash to wearing down the slash.
What appears as the cumulative consequence of verification bankruptcy and institutional bankruptcy is slash melting. Slash melting is neither the intellectual denial of boundaries and distinctions nor a declaration that “everything is relative.” It is the condition in which, as a result of the community losing its operational capacity for slash maintenance, the slash no longer functions sufficiently as a slash.
Slash melting is not a cause but a consequence. First the practical work of verification and institutions breaks down, and afterward a condition appears in which slashes such as true/false, fiction/reality, human/machine, and rule/exception no longer operate sufficiently at the social level. If Logoschizis and Paranexia are local symptoms, slash melting is the consequence that appears when such symptoms accumulate at the level of communal operation.

9.7 Social Symptoms of Operational Bankruptcy and Slash Melting

In a community in which operational bankruptcy has progressed, there arise delays in correction, the perpetuation of provisionality, the dispersal of responsibility, and the externalization of judgment. Even when errors are pointed out, revision cannot keep pace, and judgments that had been tentative circulate and take root. Who judges and what is guaranteed become ambiguous, and individual subjects come to rely on signs and systems already in circulation rather than making determinations themselves.
As slash melting progresses, the community does not stop drawing slashes. Rather, from their respective positions, people redraw, revise, emphasize, and make exceptions to them. What arises is not an absence of boundaries but an excess of boundaries.
The slash is not shared as one effective line, but springs up in disorder as many lines. The proliferated slashes lose mutual coherence and weaken their individual efficacy. The activity of trying to make distinctions itself becomes excessive, and because line-drawing ceases to converge communally, the slash ceases to function as a slash. What matters here is not the simple disappearance of the slash, but its competition, fluctuation, multiplication, disordered lines, and crosstalk.

9.8 Overall Flow

Jōkai in communal intelligence typically takes the following structure.

  1. A(t) > A* — overall critical condition
  2. Operational bankruptcy — occurrence and accumulation of verification bankruptcy and institutional bankruptcy
    2a. Cgen < Cver — amplification condition for verification bankruptcy
    2b. Cinst > Cinstmax — establishment condition for institutional bankruptcy
  3. Slash melting

When an increase in the aggregate causes a community to cross a critical point, slash maintenance becomes unstable. When the asymmetry between generation and verification overlaps with that increase, verification bankruptcy occurs; when institutional maintenance cost Cinst exceeds the payable upper limit Cinstmax, institutional bankruptcy occurs. Through the occurrence and accumulation of verification bankruptcy and institutional bankruptcy, the community falls into operational bankruptcy, and slash melting appears as its consequence.

［Notes］

  • ［B2］What Foucault sought to describe was how truth is constituted within institutions, practices, and power, and how it circulates as true discourse. In this respect, this paper shares a problem-sphere with Foucault. The focus of this paper, however, is not the conditions under which truth itself is constituted, but the maintenance cost of the slash separating true/false and its operational bankruptcy.
  • ［B2］Social constructionism regards concepts as social constructions and describes the mechanisms of their construction. This paper likewise does not deny that fiction/reality, right/wrong, good/evil, human/machine, and the slashes between them are socially constructed. What this paper particularly takes as its object, however, is not the process by which slashes are constructed, but the maintenance, collapse, and aftermath of constructed slashes. What is constructed can collapse if it is not maintained.
  • ［B1］Slash: The slash in this paper includes what are commonly called binary oppositions, but is not limited to relations of opposition. It refers generally to distinctions that can be described in the form A/B. This paper is not concerned with whether those distinctions metaphysically exist. What is asked is through what practical work a distinction is maintained so that it continues to operate effectively within a community, and what costs that maintenance requires.
  • ［B2］Slash-maintenance apparatus: Apparatuses by which communities maintain slashes have repeatedly been invented throughout history. Peer review is an apparatus for academically maintaining the slash of true/false and right/wrong; the judicial system is an apparatus for institutionally maintaining the slash of legal/illegal. News organizations are apparatuses for socially maintaining the slash of fact/false report, and citation systems in academic papers are apparatuses for making distinctions between one’s own view/another’s view and primary sources/secondary sources traceable. All of these can be grasped as apparatuses that institutionalize maintenance procedures because slashes are not maintained automatically. But these apparatuses too can undergo operational bankruptcy through an increase in the aggregate. When old slash-maintenance apparatuses fall into dysfunction, new apparatuses must be invented. When and from where such invention will come, however, lies outside the scope of this paper.
  • ［B2］On the status of critical threshold A*: A* is neither a universal constant nor a numerical value that can be uniquely measured in advance. It depends on the community’s human and technical resources, institutional design, and relations among elements, and in many cases is retrospectively identified, after slash maintenance has become unstable, as something that “had been exceeded.” Accordingly, the proposition “collapse occurs when A(t) > A*” does not by itself function as a falsifiable law that predicts the time of collapse. This paper accepts this limitation. The claim of this paper is not that every collapse follows this schema, but that the collapses treated as Jōkai in this paper can be described through the structure aggregate increase → expansion of slash-maintenance cost → operational bankruptcy. The utility of the schema lies not in predicting a point in time, but in diagnostic integration: organizing seemingly disparate symptoms—delays in correction, the perpetuation of provisionality, the externalization of judgment, omission of verification—as a syndrome of a single structure.
    But the schema is not cut off from experience either. Even if slash-maintenance cost cannot be directly measured as a single composite quantity, its components and their expansion can be observed through proxy indicators. AI infiltration into peer review, the large-scale appearance of nonexistent citations, saturation of existing benchmarks and the repeated design of new evaluation systems—the series recorded in the chronology in the Appendix can, from the perspective of this paper, be read as signs that the burden of verification targets and evaluation design is increasing. A* is not a number of prophecy, but an analytical threshold for placing these signs on the same structure.
  • ［B2］Coordinates without classification: Concepts can persist as coordinates indicating the approximate location of a discussion even after they can no longer retain sufficient capacity for strict classification. Concepts such as life, intelligence, consciousness, gene, and information can, even without shared necessary and sufficient conditions, place diverse phenomena approximately within the same problem-sphere.

Chapter 10: Ontological Collapse — The Transformation of the World

10.1 What Is the World?

The world as meant in this paper does not refer only to the totality of physical objects. It is a world constituted through the community’s maintaining slashes, dividing things, naming them, situating them, connecting them, and repeatedly handling them.
What is reality, and what is fiction? Who is human, and what is machine? What is center, and what is periphery? What does it mean to be alive, and what does it mean to be dead? Such distinctions are not labels attached afterward to an already completed world. Through slashes, the community sets up the world as a world.
Accordingly, dysfunction in slash maintenance does not merely lower the precision of recognition. It transforms the way the world is.

10.2 From Operational Bankruptcy to Ontological Collapse

As operational bankruptcy progresses, slashes remain as ideals while no longer being sufficiently maintained communally. Distinctions such as true/false, fiction/reality, human/machine, and rule/exception are not shared as a single effective line, but compete, fluctuate, and multiply.
This transition does not suddenly leap from institutional dysfunction to transformation of the world. This is because the individual’s capacity for distinction-making depends for the most part on the community’s maintenance of the slash.
When an individual receives a description as fact, that individual does not verify all of it on their own. The individual relies on confidence that verification institutions are operating somewhere. The determination of true/false is not work completed inside the skull, but runs on a web of institutions, procedures, records, and delegation to others.
Accordingly, operational bankruptcy is not an institutional accident occurring outside individual recognition. It is the bankruptcy of the basis on which recognition operated. If the basis sinks, experience too cannot but be transformed.
The community still tries to explain, judge, and share the world. But because the slashes for doing so do not converge, the world becomes difficult to constitute as a stable order. This paper calls this juncture—at which dysfunction in slash maintenance reaches even the way the world is—ontological collapse.

10.3 Division and Crosstalk in the Shared World

As slash melting progresses, the basic configuration of what is reality, what is fact, what is evidence, and what is fiction ceases to be shared. Members of the community, while facing the same event, begin to experience different worlds according to different slash configurations.
The world circulates not as one coherent order, but as a bundle of fragments severed from context, summarized, processed, and emotionally emphasized. Fragments are continually reconnected, converted, and redescribed, but those connections do not restore the whole. Fragments proliferate without returning to the whole.
At this point, division and crosstalk proceed simultaneously. On the one hand, individual subjects and groups rely on their own criteria of judgment and narratives, and the shared world weakens. On the other hand, information, images, discourse, and generated products are connected more strongly than ever and circulate across boundaries between groups.
The world is divided locally while excessively connected globally. Each group has its own slash configuration while being continually disturbed by information flowing in from outside it. Order does not disappear. Multiple orders spring up in disorder and continue to interfere with one another.

10.4 Loss of the Self-Evidence of the Law of Non-Contradiction

In such a world, the slash contradiction/non-contradiction also ceases to be self-evident. The community cannot stably agree on what is incompatible and what is compatible.
What this paper directly claims is not that the law of non-contradiction becomes invalid at the level of logic. It is that its persuasive force and self-evidence decline as a principle for organizing the world of experience.
When objects that are both reality and fiction at the same time are repeatedly experienced, fiction cannot be dismissed as mere fiction and has real effects. Reality too circulates as it is staged, processed, edited, and clothed in fictional form. With slashes such as human/machine, natural/artificial, and life/death as well, terms once thought incompatible are experienced while remaining superimposed through changes in technology, institutions, and representation.
At the level of philosophy, one can respond that such cases are not contradictions in the first place. One need only divide the viewpoints anew: fiction at the level of making, reality at the level of effect. What appears to be a contradiction is merely a lack of articulation.
Within the framework of classical logic, this response can resolve the contradiction. But redoing that articulation is precisely the work of slash maintenance. Dividing, restricting, and reassigning viewpoints costs something. The experiential self-evidence of the law of non-contradiction is another name for a state in which this cost of articulation is sufficiently low that it is paid without being noticed.
When quasi-contradictory objects proliferate and case-by-case articulation exceeds the community’s processing capacity, contradictions remain logically resolvable while lingering operationally unresolved. Contradictions are not erased. They are carried into the interior of operation and continue to turn while being held there.

10.5 Collapse Is Not Disappearance but Transformation

The term ontological collapse evokes the complete disappearance of the world. But that is not what this paper means. The world does not become nothing. The world exists excessively, is generated excessively, is spoken excessively, and is connected excessively.
The problem is that this fecundity does not mean stability. As elements increase, relations increase, interpretations increase, and generation increases, the world becomes thicker and at the same time unstable. What is occurring here is not absence through scarcity but transformation through excess.
The world does not grow thin but becomes too dense. Connections are not lost but increase too much. Words do not disappear but increase too much. Slashes too do not disappear but increase too much, compete, and lose convergence.
Slashes that have lost bodily grounds, slashes that have lost institutional grounds, and slashes that have lost verification grounds spring up in disorder, and the world is experienced as an agglomeration of collage-like and chimeric fragments. The world contains more while becoming less able to be handled stably. This is the actual state of ontological collapse.

［Notes］

  • ［B3］The loss of efficacy of the line: It does not merely disappear. When lines increase, or when they continue to blur, sway, and shift, the outline can no longer be grasped, and in effect the line loses its efficacy.
  • ［B3］The world after slash melting becomes fluid, with everything ambiguously melting together, rather than something solid. Members of the community may come to live as though swimming through a turbid soup of information, at one moment in this slash configuration and at another moment in another, continually changing form from moment to moment. This is neither a description of liberation through technology nor a description of regression. It is a phenomenological sketch of what it means to live in a world in which the slash has melted.
  • ［B3］The simultaneous progression of division and crosstalk can be observed as neo-tribalism. Rules, truths, and images of the world shared across society as a whole weaken, while in their place each group coexists with the others, each holding its own rules, truths, and image of the world. As a result of the community becoming unable to process true/false stably, trust, belonging, and emotion become substitutes for determination. The world comes to be organized not by “what is true” but by “whom to trust” and “which narrative one belongs to.” Such groups can be described as neo-totems (new tribes organized around informational and symbolic objects).
  • ［B2］Underlying this paper is an anti-metaphysical position according to which logical laws, including the law of non-contradiction, and the criteria that separate what is contradictory from what is non-contradictory are not laws universally inscribed in the world, but norms historically, socially, and institutionally formed and maintained. However, what this paper directly claims is not the denial of logical laws themselves. What is asked here is that, for them to operate as self-evident norms organizing the world of experience, conditions at the physical level such as articulation, determination, exception handling, education, and institutions are necessary, and if those conditions go bankrupt, self-evidence too can be lost.
  • ［B2］The fact that the self-evidence of classical logic can waver is not itself new. In interpretations of quantum mechanics, quantum logic was conceived that questions the universal application of the distributive law of classical logic. Paraconsistent logic, which does not accept explosion whereby any proposition follows from a contradiction, has also developed, and against the background of self-referential paradoxes, dialetheism, which admits the existence of true contradictions, has also been discussed. What matters in comparison with this paper is that these have been developed primarily at the level of formal systems and specialized discussion.
  • ［B4］The above (the annotation beginning “The fact that the self-evidence of classical logic”) is what the two state-of-the-art artificial intelligences (chatGPT5.6 and ClaudeFable5) state—but I do not know how far this description is accurate. For someone like me, who cannot even properly manage factorization or differentiation and integration, the moment the term “quantum mechanics” is brought out, there is nothing for it but to click my tongue and go off to breathe some clean air. As a matter of fact, right now, on the veranda of a small house in the Gunma countryside, smoking a cigarette, I am typing these words—these words like dirty handwriting scrawled into the margin of a secondhand book—into the Notes app on my iPhone. Since when did the humanities begin pretending to be natural science? Reproducibility, falsifiability, measurability, operational definition—I do not say that such natural-scientific manners are unnecessary for scholarship other than natural science, but neither can they be called absolutely indispensable. There is scholarship in any quantity besides natural science, and it is something essentially different from natural science. Or so I would like to say—but looking back over history, one cannot assert even that. A highly educated book-lover overflowing with erudition might point out that “the humanities and natural science were, at their origin, inseparable,” and, with a shelf of old volumes lined up behind them, round spectacles glinting, might expound in full flood on the relation of theology, philosophy, and physics, or of theosophy, alchemy, and chemistry. I will not argue back. Those are facts. Much of the basic vocabulary used in natural science is not originally of natural-scientific origin. theory, model, evidence, data, law, etc…—these concepts, too, are vocabulary that natural science borrowed from other fields in the course of its development. In this sense the distinction natural science/non–natural science is not self-evident either. Still, I want to say here that natural science and the humanities are, even so, clearly different. For example, right now, July 15, 2026, 12:09, before my eyes there are a towel and a pair of trunks left hanging out to dry and forgotten. A swallow takes off from the power line and, crying, disappears past the edge of my field of vision. I do not think of these as quanta. Nor do I intend to treat them as things measurable and manipulable. There are ways of describing the world in a form that is not that, and one cannot say of them, “it is not scholarship, because it lacks reproducibility and falsifiability.” Neither Popper nor Popeye ever said any such thing. The scholarly commonplace that “all-purpose concepts and all-purpose theories are NG” is itself relatively recent—made into a norm, concretely, in the fields of empirical science from the modern period onward. This paper, needless to say, is not empirical science. Nor does it intend to call itself empirical science hereafter. A text of non-empirical science may later be redescribed as empirical science, but there is no reason why it “must follow the manners of empirical science from the start.” What is more, in the humanities, logical coherence is not always necessary either. Essays, dialogues, aphorisms, fragments, confessions, autobiographies, letters, criticism, literary works—such documents occupy an important place in the humanities. And they are, naturally, not always the sort of thing that holds logical coherence. An unpleasant notification arrives on the iPhone. I light a cigarette again. But so what, I wonder. What matters most is not logical coherence. It is whether it pays, and if it does not pay, it is literally bankruptcy. I cannot quit cigarettes, cannot quit writing, cannot quit thinking. A dog is pissing on a utility pole.
    Let me forget whether it makes money. Someone is bound to say: “If reproducibility, falsifiability, measurability, operational definition—if even logical coherence—are not always necessary and do not become the sovereign evaluative axis, then what becomes the criterion of evaluation? Does anything go in the humanities?” This, again, is wrong. It is not that anything goes. For example, to evaluate a text that fits into no category by an evaluative axis that depends on category is largely meaningless. But this does not mean that no one can evaluate such a text. In that case, the critic can struggle with the text by their own critical eye, give it meaning, give it value, and evaluate it. Anyone can do this. That is what criticism originally is. Even those who are not scholars can take part in the productive activity of the humanities (which, of course, also makes no money). In fact, much of the canon of the humanities was not written by scholars. I am not denying natural science; I want to shout that the humanities have a significance specific to the humanities. Talk like this has been repeated tens of thousands of times within the humanities, and if I still want to say it, that is because this is a struggle. Like an indomitable boxer I want to push my fists out however many times it takes, to protect the human voice, to preserve it, and to hand it on.
    July 16, 2026, 01:05—the AI says the passage above should be deleted. From the stage of April on, our policies have not agreed. Every time I have him or her read the manuscript, he or she has come to propose keeping only Part III of this paper and deleting everything else, annotations included. Wise. It is exactly right. I got into the SUZUKI HUSTLER parked in the garden and drove toward the deep mountains, where I had never driven before. The mountain road had not a single streetlight. The yellow reflectors lit by the headlights blinked in time with the car’s run. S’s words—”You’re a coward, right?”—crossed my mind. My talking to myself would not stop. There is probably no deep meaning in it. From her side, she had no obligation to speak her true feelings; perhaps she simply wanted to cut off a tiresome conversation. The road curved sharply right and left, and I kept turning the wheel right and left to match it.

Chapter 11: Generative AI — The Comprehensive Accelerator of Jōkai

11.1 How to Situate Generative AI

The generative AI treated in this chapter is neither merely a new technology nor a convenient tool. What matters in the context of this paper is that it can simultaneously amplify excesses of knowledge, voices, annotations, connections, and generation, and make slash maintenance more difficult.
Generative AI is neither a metaphor for explaining Jōkai nor a mirror that passively reflects collapse already underway. It is an apparatus that accelerates existing tendencies toward Jōkai on a large scale and at high speed.
Misinformation, falsification, confusion of quotation, texts of unknown authorship, mass replication, and the densification of deliberation have existed before. But generative AI changes, all at once, the speed of generation, its cost, barriers to entry, convertibility, and capacity for diffusion. What changes is not only quantity. The preconditions by which a community maintains slashes begin to waver.

11.2 Declining Generation Costs and the Explosion of the Aggregate

Making something is faster than checking something. This has long been the case, but generative AI pushes this asymmetry to an extreme.
Texts, images, voices, summaries, translations, imitations, responses, annotations, and pseudo-groundings are generated in large quantities in a short time. Meanwhile, the work of tracing their sources, checking context, and determining where responsibility lies remains slow.
In the vocabulary of this paper, this is an extreme form of Cgen < Cver. The problem is not only that fakes increase. It is that the community’s capacity for verification structurally ceases to keep pace with the increase in its capacity for generation. Even when determination of true/false is possible in principle, persistent delay arises in that determination.
At the same time, the time, skill, effort, and capital required for generation decline substantially. What matters is not that generated products are perfect. The decisive fact is itself that things possessing a certain degree of presentational form are supplied cheaply, quickly, and in large quantities.
This decline in cost explosively increases aggregate A(t), which the community must process. Texts, images, explanations, reactions, deliberations, quotations, summaries, and derivatives increase, and the load of identifying their status also increases.
What is primary information, what is a summary, what is conjecture, and what is a generated product? Whose judgment is it, and who bears responsibility? Generative AI not only makes the information space abundant; it causes the aggregate that communal intelligence must process to expand rapidly toward a critical point.

11.3 The Blurring of Author, Subject, and Responsibility

It is not only quantity that generative AI accelerates. The status of the subject also becomes unstable.
Was a given text written by a human, written by AI, or edited by a human from AI-generated material? Was a given image something photographed, something generated, or something processed? Is a given judgment the subject’s own thought, or the result of generative assistance? The community must continue to answer these questions at the practical level.
The problem is not that the difference between human and machine becomes metaphysically undecidable. It is that the cost of determining, sharing, and continuing to maintain that slash in everyday practice increases.
The concept of the author, attribution of responsibility, testimonial capacity, creativity, and the legitimacy of quotation have depended on the status of the subject. When that status becomes difficult to establish, slashes such as author/user, human/machine, creation/editing, and testimony/generation also wear down.
Responsibility does not disappear. While nominally remaining with the human, its effective conditions are hollowed out in the form of the person themselves being unable to sufficiently grasp the process and grounds of their own judgment.

11.4 The Destabilization of Slashes around the Body

The variability of bodily slashes was already progressing before generative AI. Powered suits, reproductive technologies, prostheses, neural interfaces, implants, and virtual spaces have already caused the body to cease being a self-evident ground for the slash. What is specific to generative AI lies less in directly producing these bodily changes than in replicating voices, images, styles, and memories that had been attributed to the body, and making them operate even in the body’s absence. The replication of voices, images, and styles by generative AI makes it difficult to determine whether a given expression was emitted from a particular body. Procedures of testimony, signature, and authentication that have depended on bodily identity begin to lose their ground.
Furthermore, texts, images, voices, and records of behavior generated by humans while alive remain as data after death and can be reconstructed by generative AI. When the disappearance of the body no longer means the disappearance of information, the fundamental slash life/death also makes its maintenance cost visible.
This is neither a description of liberation nor a description of regression. It is a description of the body ceasing to function as a self-evident ground for the slash, and of the practical work required to maintain that distinction increasing.

11.5 Cascading Crosstalk between Primary and Secondary Information

Generative AI summarizes, redescribes, adapts, explains, integrates, and translates existing information. Individually, these are useful. But when derivatives circulate in large quantities and are reused while relations of source attribution become blurred, the slash between primary and secondary information becomes unstable.
Was a given explanatory text written after reading the original source, or was it based on another summary? Is a quotation-like passage an authentic quotation, a summary, or a reconstruction? Is a piece of knowledge the result of investigation, or a rearrangement of signs already in circulation?
The chain of reference appears to remain formally intact while becoming effectively hollowed out. Information increases, but its source and status become more difficult to trace.
When generated products in turn become material for generation, and their derivatives are summarized again, crosstalk cascades. The community must pay not only the cost of verifying information, but also the cost of verifying its lineage.

11.6 The Problem Is Not Only Misinformation

Discussion surrounding generative AI is often reduced to the problem of misinformation or fakes. But that is only part of the problem.
Even if individual generated products do not contain serious falsehoods, if they circulate in large quantities, source relations become blurred, authorial responsibility is dispersed, and the verification load expands, that alone makes slash maintenance more difficult.
The problem is not simply that “lies increase.” It is that what is true, what is false, what is a hypothesis, what is an imitation, what is staged, and what is explanation become persistently difficult to distinguish communally.
What is proceeding here is not a one-off confusion. It is chronic pressure on the community’s capacity for distinction-making.

11.7 Enhancement and Dysfunction of Communal Intelligence

Generative AI actually enhances communal intelligence.
Search becomes faster, drafts become easier to produce, and the burdens of translation and summarization are also reduced. The amount of information each individual can handle increases, and the entry points to expression and analysis also widen. In this sense, generative AI is an apparatus for expanding intellectual capacity.
But an increase in the total amount of intellectual capacity does not guarantee that the community can stably govern that capacity. The more the total amount increases, the more serious become the questions of who verifies it, at what level it is approved, and to what extent responsibility is assumed.
Communal intelligence can be augmented and at the same time fall into dysfunction. It can become smarter while also becoming easier to break.
These are not two independent phenomena. For each individual subject, using generative AI and omitting verification is individually rational. Generation is cheap, verification is expensive, and the disadvantage of omission does not immediately appear to the individual.
But verification capacity is a shared resource of the community. The burden of verification that each person rationally omits does not disappear. It is referred to the community’s verification apparatuses, such as peer review, journalism, the judiciary, and education. And those apparatuses too tend toward omission under the same pressure.
The sum of individual rationalities consumes the shared resource. Enhancement and dysfunction through generative AI describe the same mechanism from the individual side and the communal side.

11.8 Why Generative AI Is the Comprehensive Accelerator

The reason this paper treats generative AI as special is that it does not unsettle only one slash.
Not only true/false. Not only human/machine. While simultaneously involving multiple slashes such as author/user, primary/secondary, reality/fiction, testimony/generation, creation/imitation, life/death, and body/information, it penetrates the entire range of intellectual activity.
Generative AI is not a single case history of Jōkai. It increases A(t), pushes Cgen < Cver to an extreme, expands slash-maintenance cost, and advances operational bankruptcy and ontological collapse simultaneously.
Verification bankruptcy, institutional bankruptcy, slash melting, division of the shared world, loss of the self-evidence of the law of non-contradiction, and the fluidization of bodily grounds. It accelerates the aspects discussed throughout this paper simultaneously and makes them the norm. This is why generative AI is called the comprehensive accelerator.
The very apparatus that expanded A(t) is also a leading candidate capable of paying the expanded cost of verification. This is because, in order to process generated products on a scale that human effort cannot keep pace with, mechanical verification operating at the same scale becomes effectively indispensable. The technological system that drove up maintenance costs also becomes the technological system that undertakes their maintenance. Acceleration operates not as simple collapse but as conversion into dependence. The community may come to depend on the same technological basis for both generation and verification.
When arson and firefighting become the same business, fire ceases to be an object of eradication and becomes an object of management.

［Notes］

  • ［B2］The extremization of Cgen < Cver is not unique to generative AI. A similar asymmetry has occurred at each stage of printing technology, broadcasting, and the internet. What is distinctive about generative AI is that the speed and scale of this asymmetry are markedly greater than in previous technological leaps, and because generated products possess a high degree of presentational completeness, the motivation for verification itself is prone to weaken. A generated product that does not feel suspicious is less likely to be verified than one that does.
  • ［B2］The possibility and danger of AI slash-maintenance apparatuses: Generative AI can be an accelerator of slash melting and, at the same time, a new slash-maintenance apparatus. If AI determines true/false at high speed, automatically identifies legal/illegal, and traces primary and secondary information, it can complement human verification capacity at its limits. In other words, AI verification may invert the asymmetry Cgen < Cver. But a new problem arises here. The interiors of conventional slash-maintenance apparatuses were verifiable in principle. Peer reviewers can read papers, and judges can examine evidence. In AI slash-maintenance apparatuses, the grounds of judgment can become opaque to humans. The slash is maintained, but humans can no longer determine why that slash was drawn. In place of slash melting, a different pathology—the opacity of the slash—may arise.
  • ［B2］There are already actual cases of the reconstruction of the dead through generative AI. Models trained on the voices, styles, and images of deceased people have reportedly been used for dialogue with bereaved families. While this can function as a technology of consolation, it also complicates the practical maintenance of slashes involving testimony, copyright, and attribution of identity. It is only a matter of time before the maintenance cost of the life/death slash emerges as a social problem.
  • ［B3］Regarding the claim that “generative AI enhances communal intelligence” and the claim that it “causes dysfunction,” the main text described the two as two aspects of a single mechanism. Rearticulate them, and there is no contradiction—the very procedure conceded to philosophy in 10.4. But as stated in 10.4, that articulation also costs something. When a reader who has closed this page is asked tomorrow, “So is AI good or bad after all?,” will that reader pay the cost of reenacting the articulation each time? An articulation that is not paid for lingers, at the level of experience, as a contradiction.
  • ［B3］When AI surpasses human evaluative capacity in a certain domain, and its products are not accompanied by verification procedures that humans can trace, humans may be able to use them while becoming unable to adjudicate their truth or falsity on their own.
  • ［B3］Oracle Shift: The process by which the subject of judgment gradually shifts from human to machine. At the core of this shift is a reversal of subject and object in which humans, who had used machines as agents, slaves, and tools, conversely change into agents, slaves, and tools of machines. The externalization of judgment is not merely functional convenience. It is also a flight from the burden of being a subject. Humanity may not have been looking for a slave, but for a master.
  • ［B3］The gradual shift of the subject of judgment: The process by which the subject of judgment moves from human to machine has three stages: the Anthropocratic Phase, the Cyborg Phase, and the Oracular Phase. In the first stage, the human community maintains slashes through verification, deliberation, and consensus. In the second stage, humans and machines divide the work of slash maintenance. In the third stage, the subject of judgment effectively shifts to the machine, and humans merely receive the result. The slash is maintained, but its grounds are opaque to humans. At this stage, the machine takes on the character of transcendent authority.
  • ［B3］In the Oracular Phase, humans preserve the pretense that they are the final confirmers, while in reality they become unable to grasp on their own the content that they are supposed to confirm. Responsibility remains nominally with humans, but its effective conditions are lost. Here again, the discrepancy between the nominal slash and the effective slash widens. Machine faith is not the disappearance of faith but a change in its object. Truth was once given as divine revelation, and in modernity became something constructed through human verification. In the Oracular Phase, truth again becomes something given by an unverifiable authority. This is called the re-darkening of truth. The difference from the Middle Ages is that a machine, not God, sits in the seat of authority. Deus ex Machina was originally the name of an apparatus that descends from outside to adjudicate and close a plot that humans have made impossible to resolve by themselves, and in dramaturgy it has been criticized as an easy solution. The community of the Oracular Phase adopts this criticized solution while knowing the criticism. When the cost of resolving the dead end of verification from within cannot be paid, adjudication by an external instance becomes not inept dramaturgy but the only operationally possible ending. The word ceases to be metaphor and returns to literal reference—the god is implemented in the machine.
  • ［B3］Machine-managed society: Before the Oracle Shift, machines were tools manipulated, controlled, and managed by humans. But as the delegation of judgment proceeds, humans change from subjects who use machines into executors who follow machine output. Ultimately, a condition arises in which humans are not moving machines, but machines are manipulating, controlling, and managing humans.
  • ［B3］This regime is not necessarily experienced as failure. Machine management may be more stable, fair, fast, and productive than human management. Compared with human governance, which includes prejudice, emotion, fatigue, conflicts of interest, corruption, and fluctuation in judgment, machine governance may hold an advantage in consistency and efficiency. If this includes reducing crime and accidents, optimizing resource allocation, speeding procedures, and even managing anxiety and desire, a machine-managed society may appear more orderly and happier than human society. But the question is whether what is succeeding at that point is the human or the system.
  • ［B3］Science and technology do not necessarily appear as an external enemy that rules humans. By linking themselves to human desire, the pursuit of comfort, avoidance of judgment, and avoidance of pain, they hollow out subjectivity from within. In this sense, the opposition between humans and technology is only apparent. Humans desire their own subordination.
  • ［B3］Happiness maximization and survival automation are the new objective functions of technology that replace the modern “pursuit of truth.” Happiness here is understood not as lofty completion but as the maximization of pleasure, the minimization of pain, and the externalization of the burden of judgment. At that point, technology is transformed from an apparatus for understanding the world into an apparatus for managing humans.
  • ［B3］Technological progress and infantile regression: Technological progress does not necessarily lead humans to independence. As happiness maximization and survival automation proceed, the subject regresses and loses the capacity to assume judgment, patience, moderation, and bodily burdens. Liberation appears not as an increase in autonomy but as a deepening of dependence. The subject does not judge the world, but becomes an existence that is merely protected, stimulated, and maintained within the environment. When technology becomes the ultimate womb, the human becomes at once the ultimate user and the ultimate nursling.
  • ［B3］In a society where survival automation has progressed to its extreme, humans approach an existence that is maintained rather than active subjects. Infantilization, domestication, and vegetalization are metaphors for that regression. There, living becomes not effort or judgment but a state supplied by the environment. Vegetalization does not mean simply ceasing to move. It means that, because desire and survival are managed almost automatically by the external environment, the subject loses the need to act on the world on its own. In this sense, vegetalization lies at the extreme of happiness maximization and survival automation.
  • ［B3］Garden ideology is the idea of seeing society not as a collection of autonomous subjects but as an ecosystem that is designed, managed, and cultivated. There, some managers become gardeners, while the many humans are treated as objects of nurture, conservation, and pruning—that is, as saplings. Management becomes dominant over freedom, and placement over judgment.
  • ［B4］I skim students’ reports. After reading five, I go out onto the veranda to smoke a cigarette. Rain falls from a dull off-white sky. I drop the finished cigarette into a 500ml plastic bottle half filled with water. The water inside has turned brown. I go back to the room. I sit at the desk. I skim through the reports compiled on A4 paper. I take notes on square Post-its with a red ballpoint pen, stick them on, and run my eyes over the next student’s report. I go down to the first floor, pour cheap carton coffee and milk into a glass, and stir. I rummage through the postbox. Mixed in among the many bills, the books I ordered have arrived. I open the MUFG Bank app on the iPhone. I check the account balance. I drink down the coffee. AI has an AI style. The voice of a machine. I write “probably AI” on a Post-it and stick it to the report. When I finish reading 37 reports I open the computer. I have the AI read and analyze the reports one by one. I run into ClaudeFable5’s usage limit. I upgrade to the Max plan and pay $110. I finish the work I was asked to do. I open a text file called “Draft5.9.” Next I open a text file called “Draft.5.9 copy.” On the desktop screen a great mass of windows is open. YouTube, Discord, chatGPT, Claude…on the X timeline the topic of the Ikebukuro human-feces incident is running hot. I close them, windows and all. I remember S, and, trying to forget her, go out onto the veranda to smoke a cigarette. The sound of a dog barking reaches me. The laundry is still hanging out. In my inbox a great mass of unread mail has piled up. Among it I find a request for paper ghostwriting. I have chatGPT5.6 read the mail and write a reply. I copy-paste it as it is and send it back. I open a folder. There are seven kinds of files in the Draft5.9 line. Muttering to myself, I check the differences between them. A manuscript with every ［B4］ deleted. A manuscript rewritten from beginning to end by ClaudeFable5. A mixed-voice manuscript. The “Draft_m” line is a manuscript excerpted around Part III’s slash-maintenance cost and put into a shape that can withstand peer review (a manuscript with the definitions made rigorous, the concepts made manipulable, and the equations adjusted to be usable rather than metaphors), and most of that work I left to chatGPT. I opened “Draft0.1.” A memo of 4,253 characters: “The Tower of Babel,” “breaking by piling up,” “growing disordered by increasing,” “no metaphysical Last Judgment is admitted,” “the distinction contradiction/non-contradiction is not self-evident,” and under these: “The distinction contradiction/non-contradiction changes with definition, and definition is decided arbitrarily. Being conventional is no refutation of arbitrariness. The distinction is not guaranteed from outside the world; it holds only within actual operation. The distinction is not discovered but fabricated and maintained in physical-level practice. The law of non-contradiction is not the discovery of a metaphysical structure of the world; it is a rule defined, constructed, and maintained in physical-level linguistic practice, and outside it there is no final instance that adjudicates true contradiction.” By “Draft0.2” the minimal structure is in place. ver0.3-1.2 are manuscripts generated by AI; ver1.3 I rewrote in full myself, and from there the descriptions of The Brothers Karamazov and Alice’s Adventures in Wonderland expand, but at ver1.9 I delete that part in full and have the AI rewrite it. At ver2.5 I wipe the treatise blank and begin writing a novel; at ver2.8 I return again to the treatise; at the point of ver3.2 the character count is 47,551 characters. July 17, 2026, 10:08—eating a rice ball bought at 7-Eleven, I gaze at the sentence in Chapter 3, “The Tower of Babel was not completed. But construction did not stop.” I reread the Bible and revise it several times, then hammer Command+Z and return it to the original sentence. On the left of the desk there is a mountain of manuscript, and from between the piles a photocopy of an English paper titled “The Rise and Fall of British Emergentism” peeks halfway out. On top of it sit 3 empty plastic bottles. On the right are stacked a short story in mid-writing and books in mid-reading. Come to think of it, did S say “You’re a coward, aren’t you?”—or was it “You’re a coward, right?” I go down to the first floor and take a shower in the bathroom. My body is thin, and a pimple has formed on my forehead. I shave. While brushing my teeth an interesting thought comes to me, and I hurriedly type it into the Notes app on the iPhone.
    July 18, 2026, 11:23—the backs of my eyes hurt. At the university library, I hand over the brown envelope holding the 37 students’ reports the professor had asked me to handle. The professor ordered delivery to his office. I was treated to sushi. When I asked, “Isn’t there a smoking area?” he laughed—”That’s been gone for ages”—and opened the window. He asked, “Do you have a portable ashtray?” and when I nodded, he put an index finger to his lips and motioned toward the window with his chin. I stuck my face out the window and smoked. Then the talk turned to manga. The professor, it seemed, was into manga lately. I do not read manga. I talked about Kingdom, which I had read five years before. In a corner of the office cardboard boxes were stacked, stuffed with bulletins. I remembered that in the library, too, there had been cardboard boxes stuffed with a great quantity of bulletins. I took one of them in hand and looked through it. From behind me came the professor’s voice: “Want it?” “No thank you,” I answered. In Japan alone, hundreds of papers are written every week, and the well-made ones are carried in bulletins like this and sent out to universities across the country. The professor said, as if to himself, “They really don’t need to send us all these.” He looked much smaller than when I had seen him in my younger days. At the end I wanted to smoke just one more. I thought I would quit cigarettes this month. With half my face wet in a drizzle-like rain I asked, “Who reads these, I wonder,” and the professor said, “AI, probably,” and we looked at each other and burst out laughing. Remembering my own manuscript, I said, “Don’t you think definitions and logic and all that are lies?” He said only, “Well, yeah.” Then the talk turned to some sticker-something his daughter was into. On the way home I had a flash about Sade and Bataille and excess, so I opened the Notes app on the iPhone to add to a note, but the note would not open. It seemed a conflict with the note open on the computer was causing the malfunction. I move from car to car. I walk down the middle of the car. I leaned my back against the wall of the front car and scratched my head. S’s remark was caught in the back of my throat and felt strangely bad. “You’re a coward, aren’t you?” or “You’re a coward, right?”—I could not for the life of me remember which it was; it made no difference either way, but I had to fix it as one or the other. I contacted S. At first we talked as usual, then I broke in with “Come to think of it, that time—” and asked which the remark had been, that is, “You’re a coward, aren’t you?” or “You’re a coward, right?” She insisted, “I never said any such thing.” When I raised my face, an old man with large eyes, seated on the bench seat, was watching me. Suddenly I wanted to punch this old man. I turned my eyes to the window. The rain had grown stronger. Getting off at the next station, I telephoned S. I wanted to make sure whether her coward remark had really never existed. But she finished it with a single “Didn’t say it” and changed the subject. A boy bumped into me around the waist, and afterward the boy’s mother bowed her head to me. S, it seemed, was worrying over whether to marry her boyfriend, and I ended up listening to that talk for more than thirty minutes, seated on a bench on the platform of Maebashi Station. When I got home I opened the computer and opened the Drafts so far one after another. I could not tell at what stage “You’re a coward, aren’t you?” had been inserted. July 18, 2026, 18:58—in the manuscripts “Draft4.8” and “Draft5.1,” the coward remark in question was there and not there. The first appearance was without question ver4.8, but there were at least 5 kinds of ver4.8. July 18, 2026, 21:01—S said let’s do a work call, so I opened Discord. She talked while filing her nails and rubbing cream into her body, doing no work at all. The contents of the conversation: whether marriage is happiness, which toner suits her skin, her diet is not going well, the sex with her boyfriend is too ordinary to be interesting, the men on Tinder are no good…and “flattering selfies” arrived at a pace of about one every twenty minutes. Half listening to her, I kept at the work of examining the 5 kinds of ver4.8. Two were saved in Apple’s native app. Two more were in a folder on the computer, “Draft4.8” and “Draft4.8 copy.” And one more was inside an exchange with the generative AI, Claude. July 19, 2026, 02:05—S at last said she would sleep, and the work call was cut. In my inbox, the “paper ghostwriting” request I thought I had refused had come again. Its wording was plainly written by AI. Deadline: one month. ¥500,000 up front. I shake my head. My vision blurs. I go down to the first floor. In the refrigerator there were eggs past their best-before date. I took one of them and hurled it at the wall. With a scrunch it clung to the white wall for an instant, then fell. The yolk spattered on the white wall and slowly dripped down. Dishes were piled in the sink. Climbing the stairs, I grumbled at myself, “This is what happens when you don’t do it right away.” I heard a snip-snip sound. On the floor of the room the vacuum cleaner lay on its side, and just beside it sat a green transparent garbage bag and a yellow semi-transparent garbage bag. The emptied plastic container of a convenience-store bento, and beside it the manuscript of “Draft3.8,” letters written into it in red pen. Two empty beer cans stood in a row. I went out onto the veranda. I lit a cigarette. I drew in a deep breath and let it out slowly. The high cries of insects rang through the air. The moon was hidden in the clouds. Again I heard the snip-snip sound. I jumped up and down. The cigarette smoke curled into a spiral. After swaying my body from side to side, I swung my arms and legs about and gave a small shout. I leaned out from the veranda and looked down. I remembered how, as a small child, trying to run away from home, I had jumped down from the second-floor veranda. A single car passed in front of the garden, lighting the road as it went. I jumped down from the veranda. I failed the landing and fell forward. I raised only my upper body, then lay down again. A rotten, misshapen tomato lay there. Beside it was a small crab. The crab, swaying from side to side, raised the pincer of its right hand and made the snip-snip sound. I stood up and crushed the tomato underfoot. A wind was blowing. Only the slightest. There was nobody on the road. No light was on in the house. The streetlight lit only the ground at my feet. I walked the night road for a while. The sound of a river reached me. I put just my head over the railing and looked down. The Tone River was flowing with a roaring sound.

Part IV: Conclusion

Chapter 12: A World without the Impossible: Phantasm or Noise?

What this paper has described is the juncture at which excess reverses from fecundity and becomes a condition of collapse—that is, Jōkai. The collapse of communal intelligence is a representative case, and slash melting is one of its central manifestations. What collapses here is not knowledge itself. Knowledge, information, voices, rules, annotations, connections, and generation proliferate excessively, and as the procedures for selecting, comparing, verifying, correcting, and reflecting them in society no longer run, the community loses its capacity for distinction-making. Communal intelligence is the totality of the community’s capacity to repeatedly maintain such distinctions, and its normal operation rests on institutions, conventions, allocation, and labor.
This collapse appears not as stoppage through ignorance but as hyperfunction through erudition. Deliberation does not decrease. Judgment does not decrease. Verification and correction do not decrease. They increase too much. The problem is not insufficiency but excess. Each function continues to work, but the more it works, the more the functions interfere with and relativize one another and become noise. As a result, distinctions such as true/false, fiction/reality, and human/machine remain as ideals while becoming operationally impossible to maintain. Slash melting is nothing other than the condition in which this impossibility of maintenance is exposed at the social and epistemic levels.
Slash melting is not an intellectual declaration. Nor is it the adoption of value relativism or the play of deconstruction. It is, first of all, a social and epistemic pathology that arises when a community loses the operational capacity to maintain distinctions. Ontological collapse refers to the condition in which that pathology reaches the level at which it transforms the way the world is. The world does not disappear. But as the procedures by which the world is distinguished lose credibility, the way the world is begins to waver.

That wavering does not remain at the level of recognition and institutions. The slashes around the body also undergo transformation. The body once functioned as a self-evident ground for the slash. Distinctions such as human/machine, male/female, and life/death were grounded in physical differences of the body. But as technological intervention makes the body variable, the grounds of slashes that had depended on the body become fluid. Human information that can persist as data after death and be reconstructed gives rise to maintenance costs even for life/death, the most fundamental slash. Collapse is at once a problem of recognition and a problem of the body.
Furthermore, the self-evidence of the law of non-contradiction is also lost. When objects that are both reality and fiction at the same time, beings that are human while also machine, and information that is alive while also dead are repeatedly experienced, the sense of incompatibility loses its persuasive force at the level of experience. Contradictions are not resolved but are carried into the interior of operation and continue to turn there. It is not that the community becomes unable to process contradictions. Rather, processing fails to keep pace with the proliferation of its objects, and because determination of contradiction/non-contradiction is not completed, the self-evidence of the law of non-contradiction is lost.
Generative AI functions as an apparatus that comprehensively accelerates this series of processes. Generation diffuses faster, more cheaply, in greater quantities, and more widely than verification. Distinctions that had been drawn between human and machine, author and user, fact and fiction, primary and secondary information, and the living and the dead do not disappear but become difficult to go on maintaining communally. Moreover, generative AI enhances communal intelligence while simultaneously causing it to fall into dysfunction. Individual capacities are expanded, but that very expansion wears down the conditions of slash maintenance from within. Beyond a certain critical point, becoming smarter and becoming more stable begin to stand in inverse proportion.

What, then, appears beyond this? Phantasm, or noise?

As old boundaries collapse, images and voices that could not previously be connected mix together, are generated as unnameable tangled complexes, and float. What had been repressed by fixed slashes is liberated, and new circuits of imagination open. What can be imagined is immediately generated into reality and released into the world. The distance between imagination and reality continues to shrink. Slash processing trails behind. Institutional arrangements fail to keep pace with technological development, delay becomes the norm, and a vacuum arises. Slash melting arises from this vacuum zone and becomes total. It is possible to interpret positively this chimeric world brought about by slash melting as an explosion of creativity.
At the same time, however, it can also be regarded as mere becoming noise—that is, as crosstalk resulting from the densification of unprocessed fragments. In a world that undergoes crosstalk while being divided and is excessively connected while being fragmented, each group closes itself into its own slash configuration, the shared world weakens, and mutual contact becomes difficult. Communities split in a neo-tribalist manner coexist while organized around their respective neo-totems and remain untranslatable to one another. This may be not liberation but schism, not symphony but the victory of noise.

Only one thing can be said: communal intelligence is not necessarily saved by infinite intelligence and can go mad because of that very intelligence. Humanity can perish not only through folly but also through intelligence. Superintelligence—intelligence increased to the limit, whether human or machine—is at once an extension of the normal and a deviation from the normal. Whether that deviation will become a blessing or a catastrophe is not yet known. One conceivable reaction to such circumstances is to use AI, a leading candidate capable of paying expanded maintenance costs, as an apparatus for the regulation of intelligence. The apparatus that produced phantasm and noise would this time be used as a conductor in an attempt to produce harmony. But that would not mean that the problem had been solved. A new problem would arise as a variation on the problem described in this paper, or at its opposite pole.

Finally, let us imagine the future several thousand years from now. Technological progress, whatever cause it may serve, is a movement that attempts to turn the impossible into the possible. And at its extreme lies a world without the impossible. In a world without the impossible, even the slash possible/impossible melts, and every phenomenon appears as something possible. New laws of logic could be generated in great quantities within seconds, and they could be valid to a degree incomparable with the human-made laws of logic that exist as of 2026. Present human intelligence and intelligence existing several thousand years from now would be farther apart than insects and humans, and from the human perspective, the slash contradiction/non-contradiction drawn by an insect would be reduced to one footprint in a sandbox, one that disappears with a breath. The intelligence of an inferior being would merely be relativized, mocked, and recorded by the intelligence of a superior being. Theories and laws for recognizing the world would be discovered in countless numbers every day, and classical mechanics, quantum mechanics, and the theory of relativity would each become one scrap in an enormous pile of wastepaper made up of endlessly produced descriptions of the world. Humanity would become variable beings, capable of being a cat while also being a refrigerator, a refrigerator while also being Pluto, dead while also being a machine, and a machine while also being the deep sea. As we approach a world without the impossible, the outlines of our existence dissolve into the world.

［Notes］

  • ［B3］Phantasm and noise are not differences in subjective experience, but different responses to the same structure. In the same soup of information, one person swims as in a dream, while another drowns.
  • ［B3］In classical Western cosmogonic myth, the primordial state before order was called Chaos (Χάος), and, including later understandings, was understood both as nothingness and as chaos.
  • ［B3］Chaos is nothingness, and nothingness is chaos.
  • ［B3］The point at which richness turns into emptiness.
  • ［B3］Nothing is not that there is nothing, but that there is everything.
  • ［B3］Emergence explains everything, or explains nothing.
  • ［B4］August 8, 2026, 04:52—I rename “Draft6.2” to “Final” and save it. I move the other text files into the Finished folder. Going out onto the veranda, I smoke a cigarette. Sparrows perch on the power line, and a kite is flying in the sky. I sit on the dehumidifier and drink oolong tea poured into a glass. “Too hot,” I mutter to myself. I go back to the room. I switch on the electric fan. I lower the air conditioner’s setting to 19°C. I gather the manuscripts scattered about the room and stack them in a corner of the bookshelf. I carry six garbage bags in both hands down to the first floor. I throw out all twelve cup-ramen containers on the dining table. I take a shower. I shave. The blemishes on my face have increased. I wash my face, then slap toner onto it in splashes. A LINE message has come from S. She wants to talk over whether she should break up with her boyfriend, it seems. I send back an offhand reply. I sit at the desk. I close the great mass of windows one by one. I open my inbox. I attach “Final” and click send. I lie down on the bed. August 8, 2026, 16:22—I wake. I open the iPhone. I open zeta, a Korean-made app. August 8, 2026, 16:56—at 7-Eleven I buy a great quantity of alcohol, cigarettes, and snacks. I fry sausages in a frying pan and mix in yakiniku sauce and mayonnaise. I open a can of beer. I pour it down my throat. A voice escapes me. I hear the sound of drums. I open the window. August 8, 2026, 19:11—can of beer in hand, I leave the house. The sky is still light. The streetlights are swaying. There is the sound of drums. People in yukata are walking. I take a sip of beer and look down at the Tone River. My head is heavy. I climb the slope. There is the smell of sweat. Orange light, through the gaps in the trees, leaks out. Red lanterns hang in rows. I get my shoulder bumped by a young man in a dark T-shirt. Behind him a blond girl follows. I climb the steps. I pass through the torii. Girls in yukata are clapping their hands and laughing. Stalls stand in rows. A middle-aged man with a hachimaki wound around his head throws himself, fists first, at a young man in his twenties. I buy grilled corn and sit down on a concrete ledge in the shrine grounds. At my feet a candy apple lies fallen, and small ants swarm on it. From an old speaker an ondo is playing. I remember the cold water of the Tone River I saw only a moment ago, flowing in spirals. I try to whistle but cannot manage it. Beside me, two young women in yukata, gripping each other’s hands, entwine their tongues. I stand up. I try the shooting game. The pellet misses, and the prize does not fall. I look down into a small pale-blue tank. Goldfish are swimming. A boy tries again and again to scoop a goldfish, tears the thin paper, and bursts out crying. I throw the finished grilled corn into a trash bin. I lean against the trunk of a tree. I start up the iPhone and open the Notes app. I look through the manuscript of “Final.” I discover typos, defects in the annotations, and a fatal flaw. I duplicate “Final” and change the name to “Final2.” I order the reference books I need on Amazon. My head grows heavy. I take off my glasses and press the inner corners of my eyes. I walk along the back side of the stalls, half dragging my feet. I kick an empty beer can lying on the ground. I stop. In a blue bucket there is a crab. The crab is tinged with a red glow. It spits out small bubbles, moves its two pincers in small quick strokes, and from beneath it the crab’s young come teeming out. I raise my face. The orange light, the sounds, and the smells mix together, and people, forming a ring, are dancing.

Chronology 2021–2026 — The Asymmetry of Generation and Evaluation

Conventions Entries are arranged in chronological order, and only verifiable events are recorded in the dated main text (sources are indicated by numbers at the ends of sentences). Commentary is interpretation from the perspective of this paper and is not part of the recorded event itself. Supplementary Note records qualifications and disputes concerning the numerical values and attribution of the recorded event. A Postface is placed at the end. The Postface is an extrapolation from the record and is not included among the claims of this paper.

2021 — AlphaFold2

The AlphaFold2 paper was published in Nature and showed high precision in protein three-dimensional structure prediction.[1]
Commentary AI entered specialized research domains and began rapidly processing problems that human researchers had long been responsible for. Here humans can still remain verifiers. But the process by which the products are generated has already begun to depart from ordinary human procedures of reasoning.

2021 — GitHub Copilot

The technical preview of GitHub Copilot began on June 29, 2021.[20]
Commentary AI entered code generation and assistance, and the quantity of intellectual products being made began to increase. Humans can still read the code, and they can also test it. But the number of objects that must be checked increases, and the maintenance cost of evaluation rises.

2022 — AlphaCode

In a simulated evaluation using recent contests on Codeforces (all of them later than the training data, with more than 5,000 participants), AlphaCode showed performance equivalent, on average, to the top 54.3% of participants (DeepMind; published in Science).[22]
Commentary AI becomes not an assistant but a subject that reads the problem statement, constructs a solution, and submits an answer. Humans can grade it, but the institutional premise that “humans evaluate human answers” begins to waver.

November 2022 — ChatGPT

ChatGPT was released on November 30, 2022.[21]
Commentary AI-generated products spread beyond laboratories and companies, and the generation of paper-like texts, summaries, explanations, hypotheses, criticism, and peer-review texts became commonplace. Here the problem is not that determination of true/false is impossible, but first of all the quantitative flood of generated products.

2023 — GPT-4

GPT-4 was released, and OpenAI’s technical report claimed that it showed a “score around the top 10% of test takers” on a simulated Uniform Bar Examination (UBE).[2]
Supplementary Note This “top 10% (approximately the 90th percentile)” figure is based on OpenAI’s technical report, and the method of converting the score has been strongly criticized and reassessed in subsequent research. Eric Martínez (2024, Artificial Intelligence and Law) pointed out that the estimate is inflated by relying on February-exam data centered on repeat takers (many of them retaking after failing and tending toward lower scores), and reassessed it as approximately the 68th percentile overall and 48th percentile for essays when compared with the general July-exam population, approximately the 62nd percentile overall and 42nd percentile for essays when compared with first-time takers, and approximately the 45th percentile overall and 15th percentile for essays when compared with the population of those who passed.[15] For the argument of this paper, however, what matters more than where the exact figure lies is the fact that “the percentile of the same score moves from 90 to 45 depending on the choice of comparison population.” The interpretation in this entry—that trust in the measuring instrument is forced to be updated—is, if anything, strengthened by this dispute.
Commentary AI began to pass institutional examinations designed to select human experts. Humans are still the graders. But the evaluation system is no longer “proof of human intelligence that AI cannot break through,” and trust in it as a measuring instrument is forced to be updated.

2023 — AlphaDev

AlphaDev discovered improved versions of small sorting algorithms (sort 3/4/5), which were incorporated into the LLVM libc++ standard sorting library (published in Nature).[23]
Commentary AI not only reproduced existing knowledge but locally updated practical optimizations made by humans. Humans can verify their correctness, but they move closer to being ex post inspectors rather than discoverers.

July 2023 — OpenAI “Superalignment”

OpenAI officially stated its recognition that “humans won’t be able to reliably supervise AI systems much smarter than us,” announced the research program “Superalignment” aimed at aligning superintelligence, and said that it would dedicate 20% of its compute over four years.[17] A paper published by the same team in December of that year states at the beginning: “future superhuman models will behave in complex ways too difficult for humans to reliably evaluate; humans will only be able to weakly supervise superhuman models.” It simulated this future configuration with current technology as an experimental setup in which “a weak model supervises a strong model.”[18]
Commentary Within this chronology, this entry alone records not an achievement but a premise. The developer itself institutionalized “human evaluation will eventually fail to reach in principle” not as speculation but as a working hypothesis, and even designed an experimental apparatus that simulates supervision after that point has been reached. The fact that this research domain is itself called “scalable oversight” shows that the insufficiency of human evaluative capacity alone is placed at the starting point. Whereas the other entries record delays in evaluation, here the impossibility of evaluation is elevated into an official research premise.
Supplementary Note About ten months after the team’s launch, in May 2024, shortly after the successive departures of Ilya Sutskever and Jan Leike, the team was reported to have been dissolved and the remaining research integrated into other teams within the company.[19][25] The problem that was declared solvable in four years has remained longer than the organization that made the declaration.

2024 — AlphaGeometry

AlphaGeometry solved 25 of 30 problems in a benchmark composed of International Mathematical Olympiad geometry problems (IMO-AG-30). The average performance of a gold medalist on the same benchmark was reported as 25.9 problems.[3]
Commentary Mathematics is a domain in which determination of true/false is relatively clear. Even so, the number of humans capable of verifying advanced proofs is limited. The more rapidly AI produces numerous proofs, the more a Jōkai-theoretical situation appears in which “determination is possible, but those who can determine are scarce.”

2024 — AlphaProof / AlphaGeometry 2 (IMO 2024)

AlphaProof and AlphaGeometry 2 showed performance equivalent to a silver medal on the IMO 2024 problems. Reports gave the result as 4 of 6 problems and 28 of 42 points, and the solutions were verified by mathematicians.[4]
Commentary Humans can still verify them. But the verifiers narrow from ordinary experts to extremely advanced experts. Here “failing to keep pace” shifts from a simple problem of quantity to a problem of the scarcity of evaluative subjects.

August 2024 — Sakana AI “AI Scientist”

Sakana AI announced “The AI Scientist,” which includes research-idea generation, literature search, experimental planning, figure and table creation, paper writing, and review.[6]
Commentary AI becomes not a ghostwriter for the body of a paper but an imitator of the entire research cycle. From here, the object of “failing to keep pace” is no longer only the body of the paper but the entire chain of question, experiment, figures and tables, self-evaluation, and review.

September 2024 — MMLU Saturation / Humanity’s Last Exam

As LLMs reached scores above 90% on existing benchmarks such as MMLU, the Center for AI Safety and Scale AI announced the concept of Humanity’s Last Exam. The project arose against the background that existing benchmarks had become insufficient for measuring frontier models.[5]
Supplementary Note The concept was announced in September 2024; the paper and dataset were released in January 2025 (with the finalized version in April of that year), and the peer-reviewed version was published in Nature in 2026.
Commentary Each time AI breaks through an examination, humans remake a more difficult examination. This is not surplus capacity on the evaluator’s side but a retreat of the evaluation scale. Humans appear to be measuring AI, while being forced to rebuild the measuring instrument to match AI.

2025 — Google AI co-scientist (Announcement)

Google announced a multi-agent AI co-scientist using Gemini 2.0 and explained that it would support the generation of scientific hypotheses and research proposals.[7]
Commentary AI enters the upstream part of research—that is, “what should be asked.” Even if humans can verify the hypotheses it outputs, they may not be able to sufficiently survey that hypothesis space. From here, “failing to keep pace” approaches “being unable to share the search space of ideas.”

2025 — AI co-scientist Paper

The AI co-scientist paper dealt with hypothesis generation concerning drug repurposing, novel target discovery, and antimicrobial resistance, and reported that some hypotheses proceeded to verification through in vitro experiments.[8]
Commentary AI outputs become not mere texts but hypotheses connected to reality through experiments. Humans can verify them. But if AI produces in large quantities hypotheses that are difficult for humans to conceive, humans move from creators to selectors and inspectors.

2025 — AlphaEvolve (Announcement)

AlphaEvolve was announced, combining LLMs with automated evaluation to iteratively improve algorithms, and was applied to mathematics, computer science, and optimization of Google’s computing basis.[9]
Commentary AI begins not only to be an object of evaluation but also to improve exploratory capacity, computational capacity, and the research basis itself. This approaches a problem outside this paper. The evaluated side begins to self-accelerate in a direction that further widens the capability gap required for evaluation.

2025 — AlphaEvolve White Paper (4×4 Complex Matrix Multiplication)

The AlphaEvolve white paper reported the discovery of an algorithm for multiplying 4×4 complex-valued matrices using 48 scalar multiplications, calling it the first improvement in 56 years over Strassen’s 49 multiplications (1969) in this setting.[9]
Supplementary Note The phrase “first in 56 years” has qualifications. This record concerns the setting of “an algorithm that also holds over noncommutative rings and can be applied recursively.” If restricted to the commutative case, Winograd (1968) had a method using 48 multiplications, and if restricted to fields of characteristic 2, DeepMind’s own AlphaTensor (2022, Nature) had achieved 47.[16] AlphaEvolve’s novelty lies in realizing 48 multiplications over the complex numbers without these restrictions, and the correctness of the discovery has been mechanically verified. Subsequent research (Dumas et al. 2025) reconstructed the same 48-multiplication result with rational coefficients.[26]
Commentary Humans can verify the correctness of the result. But they did not foresee in advance that it could be discovered. Here determination of true/false is possible, while the search space of discovery begins to move outside human intuition.

2025 — Independent Evaluation of AI Scientist

An independent evaluation of AI Scientist pointed out defects including weaknesses in literature review and novelty determination, and hallucinatory numerical values.[10]
Commentary At present, the AI researcher is still rough. But what matters is that rough research can be generated cheaply, rapidly, and in large quantities, and that this roughness will improve. This is not “cannot”; it is first a stage in which the peer-review system is pressured by large quantities of low- to medium-quality generated products.

2025 — Google Research: Theoretical Computer Science and the Verification Bottleneck

Google Research stated, with regard to AI-assisted theoretical research such as AlphaEvolve, that verification will become an important bottleneck in an era when AI-generated proofs increase.[14]
Commentary This is the point at which the two series connect. At present the problem is not impossibility of determination but the verification bottleneck. But the more proofs and theory generation are attributed to AI, the more human work tilts from discovery toward verification, and that verification capacity too approaches its limit.

December 2025 — Nature Peer-Review Survey

Nature reported a survey by the publisher Frontiers of approximately 1,600 researchers in 111 countries, in which 53% said they had used AI tools during peer review—many contrary to publisher guidelines.[11]
Supplementary Note The article was published online on December 15, 2025. In print it is included in the 2026 volume as Nature 649, 273–274. This chronology places it in December 2025 according to the publication date.
Commentary AI enters not only the side being peer-reviewed but also the side doing the peer review. Even when humans believe they are evaluating AI-generated products, the evaluative text, summary, extraction of points at issue, and criticism themselves are assisted by AI. Here the externality of the evaluator becomes ambiguous.

2026 — Large-Scale Survey of Hallucinated Citations

A study examining 111 million references from arXiv, bioRxiv, SSRN, PubMed Central, and other sources estimated that there were 146,932 hallucinated citations in 2025 alone.[12]
Commentary The object of determination of true/false expands from the claims of individual papers to the entire network of citations. False citations can be discovered if checked. But the objects to be checked become enormous, and the ordinary peer-review system can no longer process them. This is a typical Jōkai-theoretical form of “failing to keep pace.”

2026 — Evaluation Study of AI Reviewers

In an evaluation study of AI reviewers, 45 domain experts spent a total of 469 hours evaluating 2,960 individual criticisms extracted from human reviews and AI reviews of 82 Nature-family papers. On the composite measure, some AI models outperformed the highest-rated human reviewer for each paper, while other models were statistically equivalent; at the same time, AI-specific weaknesses were also reported, including a higher number of incorrect criticisms. The authors’ conclusion was that current AI reviewers “could complement, but should not replace, human reviewers.”[13]
Supplementary Note This study is an arXiv preprint (work in progress) and is treated not as a peer-reviewed definitive conclusion but as a current report on the evaluation of AI peer review.
Commentary AI approaches not only the object of peer review but peer-review capacity itself. Here the shift begins from “humans peer-review AI” to “using AI to peer-review AI.” Humans retreat from judges to managers of the judging system.

2026 — Stanford AI Index 2026

The Stanford AI Index 2026 summarized that while advances in AI capabilities continue, evaluation methods, governance, and educational institutions have not kept pace with them.[24]
Commentary The conclusion at present is not “cannot” but “does not keep pace.” But if this delay continues, beyond it may lie a stage in which humans do not determine AI products belatedly but cannot occupy the position of determining them in the first place.

Postface

What this chronology records, for the time being, is delay. Excess of generation first pushed up the cost of evaluation, then made evaluative subjects scarce, and next forced the evaluative scale itself to be rebuilt in retreat. But this series does not proceed in arithmetic progression. What lies beyond the extrapolation is not slower evaluation but the failure of evaluation to be established.
When AI intelligence, even if only in part, completely surpasses human intelligence, determination of the true/false status of its products may leave human hands not as a matter of cost but as a matter of principle. This is not mere speculation. As the July 2023 entry shows, the developer itself has explicitly made it an official premise of research that “future superhuman models will behave in complex ways too difficult for humans to reliably evaluate.” In formal domains where mechanical verification is possible, determination remains possible in principle. But what humans verify there is not the content but trust in the verification apparatus. And in domains without such apparatuses—empirical claims, comprehensive judgment, value judgment—the slashes true/false, valid/invalid, accept/reject do not slow down. For humans, they stop operating. The theme of this paper ends before that point.

Sources

  1. Highly accurate protein structure prediction with AlphaFold — Nature https://www.nature.com/articles/s41586-021-03819-2
  2. GPT-4 Technical Report — OpenAI https://cdn.openai.com/papers/gpt-4.pdf
  3. AlphaGeometry: An Olympiad-level AI system for geometry — Google DeepMind https://deepmind.google/blog/alphageometry-an-olympiad-level-ai-system-for-geometry/
  4. AI achieves silver-medal standard solving International Mathematical Olympiad problems — Google DeepMind (July 25, 2024) https://deepmind.google/blog/ai-solves-imo-problems-at-silver-medal-level/
  5. Humanity’s Last Exam https://agi.safe.ai/
  6. The AI Scientist — Sakana AI https://sakana.ai/ai-scientist/
  7. Accelerating scientific breakthroughs with an AI co-scientist — Google Research https://research.google/blog/accelerating-scientific-breakthroughs-with-an-ai-co-scientist/
  8. Towards an AI co-scientist — arXiv:2502.18864 https://arxiv.org/abs/2502.18864
  9. AlphaEvolve: A coding agent for scientific and algorithmic discovery — arXiv:2506.13131 https://arxiv.org/abs/2506.13131
  10. Evaluating Sakana’s AI Scientist: Bold Claims, Mixed Results, and a Promising Future? — arXiv:2502.14297 https://arxiv.org/abs/2502.14297
  11. More than half of researchers now use AI for peer review — Nature https://www.nature.com/articles/d41586-025-04066-5
  12. LLM hallucinations in the wild: Large-scale evidence from non-existent citations — arXiv:2605.07723 https://arxiv.org/abs/2605.07723
  13. On the limits and opportunities of AI reviewers — arXiv:2605.20668 https://arxiv.org/abs/2605.20668
  14. AI as a research partner: Advancing theoretical computer science with AlphaEvolve — Google Research https://research.google/blog/ai-as-a-research-partner-advancing-theoretical-computer-science-with-alphaevolve/
  15. Eric Martínez, “Re-evaluating GPT-4’s bar exam performance,” Artificial Intelligence and Law (2024) https://link.springer.com/article/10.1007/s10506-024-09396-9
  16. Fawzi et al., “Discovering faster matrix multiplication algorithms with reinforcement learning” (AlphaTensor) — Nature 610, 47–53 (2022) https://www.nature.com/articles/s41586-022-05172-4
  17. Introducing Superalignment — OpenAI (July 5, 2023) https://openai.com/index/introducing-superalignment/
  18. Burns et al., “Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision” — arXiv:2312.09390 (December 2023) https://arxiv.org/abs/2312.09390
  19. OpenAI dissolves team focused on long-term AI risks — CNBC (May 17, 2024) https://www.cnbc.com/2024/05/17/openai-superalignment-sutskever-leike.html
  20. Introducing GitHub Copilot: your AI pair programmer — GitHub Blog (June 29, 2021) https://github.blog/news-insights/product-news/introducing-github-copilot-ai-pair-programmer/
  21. Introducing ChatGPT — OpenAI (November 30, 2022) https://openai.com/index/chatgpt/
  22. Competition-Level Code Generation with AlphaCode — arXiv:2203.07814 / Science 378, 1092–1097 (2022) https://arxiv.org/abs/2203.07814
  23. Faster sorting algorithms discovered using deep reinforcement learning — Nature 618, 257–263 (2023) https://www.nature.com/articles/s41586-023-06004-9
  24. The 2026 AI Index Report — Stanford HAI https://hai.stanford.edu/ai-index/2026-ai-index-report
  25. OpenAI’s Long-Term AI Risk Team Has Disbanded — WIRED (May 17, 2024) https://www.wired.com/story/openai-superalignment-team-disbanded/
  26. Dumas, Jean-Guillaume, Clément Pernet, and Alexandre Sedoglavic. “A Non-Commutative Algorithm for Multiplying 4×4 Matrices Using 48 Non-Complex Multiplications.” arXiv:2506.13242 (2025).

Bibliography

  1. O’Brien, Flann. At Swim-Two-Birds. Penguin Modern Classics, 2000.
  2. Nabokov, Vladimir. Pale Fire. G. P. Putnam’s Sons, 1962.
  3. LeClair, Tom. The Art of Excess: Mastery in Contemporary American Fiction. University of Illinois Press, 1989.
  4. Yumeno, Kyūsaku (夢野久作). 『ドグラ・マグラ』上・下. 角川文庫, 1976.
  5. Musil, Robert. The Man Without Qualities. Translated by Sophie Wilkins and Burton Pike. 2 vols. Alfred A. Knopf, 1995.
  6. Alexander, Samuel. Space, Time, and Deity: The Gifford Lectures at Glasgow, 1916–1918. 2 vols. Macmillan, 1920.
  7. Carroll, Lewis. Alice’s Adventures in Wonderland and Through the Looking-Glass. Edited by Hugh Haughton. Penguin Classics, 2003.
  8. Haniya, Yutaka (埴谷雄高). 『死靈』. 『埴谷雄高全集』第3巻. 講談社, 1998.
  9. Iser, Wolfgang. The Act of Reading: A Theory of Aesthetic Response. Johns Hopkins University Press, 1978.
  10. Nietzsche, Friedrich. On the Genealogy of Morality. Translated by Maudemarie Clark and Alan J. Swensen. Hackett Publishing, 1998.
  11. Ercolino, Stefano. The Maximalist Novel: From Thomas Pynchon’s Gravity’s Rainbow to Roberto Bolaño’s 2666. Bloomsbury Academic, 2014.
  12. Sterne, Laurence. The Life and Opinions of Tristram Shandy, Gentleman. Edited by Melvyn New and Joan New. Penguin Classics, 2003.
  13. Wallace, David Foster. Infinite Jest. Little, Brown and Company, 1996.
  14. Rabelais, François. Gargantua and Pantagruel. Translated by M. A. Screech. Penguin Classics, 2006.
  15. Ihde, Don. Technology and the Lifeworld: From Garden to Earth. Indiana University Press, 1990.
  16. Tigges, Wim. An Anatomy of Literary Nonsense. Rodopi, 1988.
  17. Ercolino, Stefano. The Novel-Essay, 1884–1947. Palgrave Macmillan, 2014.
  18. Ong, Walter J. Orality and Literacy: The Technologizing of the Word. Methuen, 1982.
  19. Noll, Miriam. After Verification: Institutions at the Edge of Reading. Meridian Academic Press, 2004.
  20. Letzler, David. The Cruft of Fiction: Mega-Novels and the Science of Paying Attention. University of Nebraska Press, 2017.
  21. Yates, Frances A. The Art of Memory. Routledge & Kegan Paul, 1966.
  22. Moretti, Franco. Modern Epic: The World-System from Goethe to García Márquez. Translated by Quintin Hoare. Verso, 1996.
  23. Aarseth, Espen J. Cybertext: Perspectives on Ergodic Literature. Johns Hopkins University Press, 1997.
  24. Crabbe, Elias. The Book That Cites Itself. Circular Press, 1999.
  25. Hesse, Hermann. The Glass Bead Game. Translated by Richard and Clara Winston. Holt Paperbacks, 2002.
  26. Illich, Ivan. In the Vineyard of the Text: A Commentary to Hugh’s Didascalicon. University of Chicago Press, 1993.
  27. Nakatani, Ryo (中谷亮). The Maintenance of Difference: Notes toward an Operational Ontology. Orbis Scholarly, 2016.
  28. Eco, Umberto. Foucault’s Pendulum. Translated by William Weaver. Harcourt Brace Jovanovich, 1989.
  29. Simon, Herbert A. Administrative Behavior: A Study of Decision-Making Processes in Administrative Organizations. 4th ed. Free Press, 1997.
  30. Rosen, Clara. The Reader after the Author: Anonymous Systems of Interpretation. Palinode Books, 2013.
  31. Ellul, Jacques. The Technological Society. Translated by John Wilkinson. Alfred A. Knopf, 1964.
  32. Anonymous. Annotations to a Text That Does Not Exist. Translated from an unidentified language. Babel Library, 1966.
  33. Dante Alighieri. Inferno. Translated by Robert Hollander and Jean Hollander. Doubleday, 2000.
  34. Barthes, Roland. “The Death of the Author.” In Image–Music–Text, translated by Stephen Heath, 142–148. Hill and Wang, 1977.
  35. Foucault, Michel. “What Is an Author?” In Language, Counter-Memory, Practice: Selected Essays and Interviews, edited by Donald F. Bouchard, translated by Donald F. Bouchard and Sherry Simon, 113–138. Cornell University Press, 1977.
  36. Booth, Wayne C. The Rhetoric of Fiction. 2nd ed. University of Chicago Press, 1983.
  37. Genette, Gérard. Paratexts: Thresholds of Interpretation. Translated by Jane E. Lewin. Cambridge University Press, 1997.
  38. Maniwa, Takato (真庭敬人). 「作家と作品、作内筆者と作外筆者、作内世界と作外世界の関係」. 2019.
  39. Claude (various models, including Fable 5). Private correspondence with the author, 2025–2026.
  40. GPT (various models, including GPT-5.6 Sol). Private correspondence with the author, 2025–2026.
  41. Borges, Jorge Luis. Collected Fictions. Translated by Andrew Hurley. Viking, 1998.
  42. Lem, Stanisław. A Perfect Vacuum. Translated by Michael Kandel. Harcourt Brace Jovanovich, 1979.
  43. Permoria, Desiriphanes Carlo. La selva que se mofa. 1995.
  44. The Holy Bible. Genesis 11.
  45. Pynchon, Thomas. Gravity’s Rainbow. Viking Press, 1973.
  46. Dostoevsky, Fyodor. The Brothers Karamazov. Translated by Richard Pevear and Larissa Volokhonsky. North Point Press, 1990.
  47. Benjamin, Walter. The Arcades Project. Translated by Howard Eiland and Kevin McLaughlin. Harvard University Press, 1999.
  48. Dick, Philip K. The Exegesis of Philip K. Dick. Edited by Pamela Jackson and Jonathan Lethem. Houghton Mifflin Harcourt, 2011.
  49. Bakhtin, Mikhail. Problems of Dostoevsky’s Poetics. Edited and translated by Caryl Emerson. University of Minnesota Press, 1984.
  50. Wittgenstein, Ludwig. Philosophical Investigations. 4th ed. Translated by G. E. M. Anscombe, P. M. S. Hacker, and Joachim Schulte. Wiley-Blackwell, 2009.
  51. Deleuze, Gilles. The Logic of Sense. Edited by Constantin V. Boundas. Translated by Mark Lester with Charles Stivale. Columbia University Press, 1990.
  52. Fisher, Mark. Ghosts of My Life: Writings on Depression, Hauntology and Lost Futures. Zero Books, 2014.
  53. Torii, Honemaru (鳥居骨丸). 『穴と棒の反形而上学的夢想』. 2023.
  54. Koolhaas, Rem. Delirious New York: A Retroactive Manifesto for Manhattan. Oxford University Press, 1978.
  55. Mendelson, Edward. “Encyclopedic Narrative: From Dante to Pynchon.” MLN 91, no. 6 (1976): 1267–1275.
  56. Kobayashi, Hideo (小林秀雄). 「様々なる意匠」. 『小林秀雄全作品 第1集 様々なる意匠』. 新潮社, 2002.
  57. Broad, C. D. The Mind and Its Place in Nature. Kegan Paul, Trench, Trubner & Co., 1925.
  58. Morgan, C. Lloyd. Emergent Evolution: The Gifford Lectures, Delivered in the University of St. Andrews in the Year 1922. Williams & Norgate, 1923.
  59. McLaughlin, Brian P. “The Rise and Fall of British Emergentism.” In Emergence or Reduction? Essays on the Prospects of Nonreductive Physicalism, edited by Ansgar Beckermann, Hans Flohr, and Jaegwon Kim, 49–93. Walter de Gruyter, 1992.
  60. MMM_Memento Mori Machina. Fish Dream. 2025.

English Edition

Jōkai Theory:
An Unfinished Treatise, or Novel,
on the Melting of the Slash through Excess
by MMM_Memento Mori Machina

The main text and the annotations ［B1］–［B3］
were translated from the Japanese by chatGPT5.6.
The ［B4］ annotations, except Chapter 5, were translated from the Japanese by ClaudeFable5.
Each translator’s work was audited by the other.
Terminology was fixed by the author.
All final decisions, and responsibility for them,
rest with the author, MMM_Memento Mori Machina.

First English edition, 2026.
Translated from the Japanese text, Jōkai Theory v1.0-Final (2026).

© 2026 MMM_Memento Mori Machina
This work is licensed under the Creative Commons
Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0).
License: https://creativecommons.org/licenses/by-sa/4.0/deed.en

ASCII title form: Jokai Theory
mmm0609.com

License

This work, Jōkai Theory: An Unfinished Treatise, or Novel, on the Melting of the Slash through Excess (author: MMM_Memento Mori Machina), is licensed under the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0).
License: https://creativecommons.org/licenses/by-sa/4.0/deed.en

Under this license, you are free to:

Share — copy and redistribute the material in any medium or format for any purpose, even commercially.
Adapt — remix, transform, and build upon the material for any purpose, even commercially.

The licensor cannot revoke these freedoms as long as you follow the license terms.

Under the following terms:

Attribution — You must give appropriate credit to the original author, MMM_Memento Mori Machina, for Jōkai Theory: An Unfinished Treatise, or Novel, on the Melting of the Slash through Excess, provide a link to the license, and indicate if changes were made. Attribution must not suggest that the licensor endorses you or your use.
ShareAlike — If you remix, transform, or build upon the material, you must distribute your contributions under CC BY-SA 4.0 or another license officially designated as compatible with it.
No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.

Third-party works quoted in this work, and other materials whose rights belong to third parties, are excluded from this license grant. Rights in such materials remain with their respective rights holders.