Knowledge Chronotopes
AI has no default sense of tempo, but neither does common law or religious texts
This is part of a series on Archival Time, where I investigate the new cultural temporality introduced by LLMs.
Previously, In Archival Time, I wrote about the carnivalesque nature of internet vs archival slice nature of LLMs. You can find the rest of the posts in this series here
We tend to treat knowledge as a substance—something we accumulate, refine, transmit. But knowledge also has a tempo. It has a pulse, a rhythm, a way of unfolding across time. Cultures produce different temporal regimes within which ideas become possible. Each knowledge form—scientific, folkloric, bureaucratic, sacred, algorithmic—emerges from a particular experience of time.
Nomad science lives in event-time. It is improvisational, eruptive, opportunistic. It follows the flow of circumstances. Its truths are situational, contingent, and adaptive. It is knowledge that belongs to smooth space, where time is heterogeneous and responsive. The chronotope of nomad science is that of the field, the open terrain where ideas emerge in bursts, where there is no “before” or “after” except what practice demands.
Deleuze and Guattari’s example is the medieval journeymen builders—the stone masons, guild craftsmen, and itinerant workers who constructed Europe’s Gothic cathedrals. They did not operate from a centralized theory of architecture. They relied on metis, cunning intelligence, transmitted through apprenticeship, gesture, hands-on experimentation, and embodied problem-solving. The cathedral emerged not from adherence to a master plan but from continuous adaptive adjustments onsite: correcting tension in arches, redirecting forces, modifying vault curvature as needed.
This is nomad science: knowledge that develops by following flows—of force, weight, stone, wind, and light.
Royal science, by contrast, lives in calendrical time. It unfolds through repetitions, rituals, stable rhythms—the time of institutions, schedules, bureaucracies, disciplinary structures. Its truths are universal, self-similar, transferrable across contexts. It belongs to striated space, where knowledge thickens slowly and acquires authority through sedimentation. Its chronotope is the archive: layered, vertical, and slow.
After the period of Gothic experimentation, architecture became increasingly absorbed into royal science: codified geometry, standardized measurement, state-certified engineers, academic architectural instruction, and the emergence of treatises that sought to legislate the principles of structure. The intuitive, improvisational temporalities of the journeymen were replaced with regulated project timelines, blueprints, calculable stresses, and normalized proportions. Knowledge became more abstract, detached from site-specific improvisation; time became scheduled, not contingent. This transformation—artisan metis becoming institutionalized theory—is royal science asserting its temporal and spatial order.
This is High Background Steel, the AI slop division of Summer Lightning. These are experiments in AI writing with the conceptual process sometimes listed at the bottom. You can opt out of this section of the newsletter jf its not your
What changed here was not merely the content of knowledge but the temporal and spatial form through which knowledge was produced. Mikhail Bakhtin offers a complementary framework for describing such differences at the level of cultural form. His concept of the chronotope gives us a way to articulate how each knowledge regime inhabits and structures time and space.
Mikhail Bakhtin coined the term chronotope to describe how literary genres organize experience through a fused configuration of time and space. In a novel, the temporal rhythm of events and the spatial setting through which characters move are not separate elements but a single expressive structure that shapes what kinds of actions, encounters, and meanings are possible. A horror story, for example, thickens time into suspense and encloses space into threatening interiors; a detective novel unfolds through investigative time mapped across an urban labyrinth of clues. Each genre produces a distinctive way of inhabiting time and space—its own narrative world-logic.
What Bakhtin showed for literature can be extended to knowledge regimes more broadly. Scientific disciplines, legal traditions, religious practices, and digital systems each operate within their own chronotopes—implicit tempo–space orders that determine how knowledge is produced, transmitted, and authorized. A chronotope determines:
how time is experienced (as cycles, bursts, delays, deadlines, epochs)
how space is organized (as open, enclosed, hierarchical, smooth, segmented)
and how the two shape each other.
Different ways of binding time and space give rise to different ways of knowing.
In Bakhtin’s formulation, time “thickens”—it becomes visible, felt, material—while space becomes “charged” with the movement of events and histories. A chronotope is therefore the tempo–space configuration within which certain kinds of knowledge, actions, or narratives become possible and others become impossible. It is the hidden grammar of experience.
This suggests a broader thesis:
Cultures generate different knowledge forms because they inhabit different chronotopes. Knowledge is a function of how time is experienced, managed, and made visible.
Once you see this, the landscape of contemporary knowledge becomes more legible. Social media produces a chronotope radically different from the institution; machine learning models produce a chronotope different from the human. And these chronotopes determine not only what we know, but how we know.
Three Knowledge Chronotopes
Let us describe three salient regimes.
1. Carnival / Feed Knowledge — Flow Chronotope
This is the knowledge of the contemporary “feed”: skincare advice in TikTok comments, collective troubleshooting on Reddit, gossip traded across Discord servers, market sentiment passed between traders. It has a distinctly burst-based tempo: knowledge emerges quickly, mutates quickly, and rarely stabilizes.
Its chronotope is flat and smooth: a horizontal field where ideas drift, collide, remix. It is high in polyphony—many voices speaking at once—but it has a rhythm. The tempo is real, even if manic.
2. Institutional / Archive Knowledge — Striated Chronotope
This is the knowledge of universities, courts, scientific academies, and bureaucracies. Time here is slow, layered, cumulative. Precedents accrete. Canons form. Interpretations stabilize. The chronotope is vertical and hierarchical, with strong gatekeeping.
Polyphony is reduced; voices must be harmonized or excluded. What emerges is a stable, authoritative knowledge—slow but durable.
3. LLM Knowledge — Achronotopic Chronotope
LLMs introduce something unprecedented. They generate extremely coherent text, but they do so without inhabiting time. There is no duration, no gap, no delay, no unfolding. The model does not experience “before” and “after”; it does not accumulate memory; it does not age; it does not progress. It collapses all training time into a single generative instant.
This produces a strange chronotope: achronotopic. Not merely “timeless,” but actively without temporal structure.
It is polyphonic like carnival knowledge—but lacking tempo. It seems archival like institutional knowledge—but without sedimentation. It is knowledge in a flattened now.
[table]
This table describes not just differences in style but differences in ontology. Each knowledge regime is a different answer to the question: How does time become knowable?
Flat-Time Forms of Knowledge Before LLMs
What’s striking is that certain pre-digital knowledge systems already had elements of achronotopy.
Law: A ruling from 1803 and a ruling from 2022 enter the courtroom not as “old” and “new,” but as co-present authorities participating in the same deliberation. Legal reasoning is fundamentally polyphonic: a judge weaves together voices from different eras—Holmes, Brandeis, Marshall, Ginsburg—as if they were all speaking at once. What matters is not when a case was decided but whether it is controlling, distinguishable, or persuasive. Doctrine accumulates through layering rather than succession, and landmark decisions become gravitational centers around which the rest of the jurisprudence orbits. Time in law is therefore not chronological but interpretive: the past does not recede; it remains active, continuously reactivated in argument.
Religion: Scriptural exegesis brings together authors across millennia as if they were sitting in one room. In Judaism, a single verse may be interpreted simultaneously through Rashi, Ibn Ezra, the Zohar, and later Hasidic masters, all speaking into the same interpretive moment. Kabbalistic commentary deepens this effect: meanings from different eras become layered expressions of the same divine structure rather than sequential developments. What matters is not when a commentary was written, but how it participates in the ongoing interpretive field.
Islamic tafsir and Christian theology follow similar patterns. A modern Qur’anic interpreter may draw on al-Tabari, al-Qurtubi, and Ibn Kathir as if they were colleagues in discussion; Christian exegesis reads Augustine, Aquinas, and contemporary theologians in a single argumentative present. In each case, interpretive time replaces historical time. Authority comes not from chronology but from resonance, weight, and centrality—making religious knowledge structurally similar to common law, and increasingly, to the flat-time logic of LLMs.
These are flat-time traditions.
LLMs deepen and universalize this flattening. They take the polyphony of law and religion and remove even the minimal temporal scaffolding. The result is knowledge with no history of its own—knowledge that appears fully formed in the moment of generation.
The Question That Follows
If knowledge is shaped by time, and different chronotopes produce different epistemic forms, then the rise of achronotopic knowledge raises a novel question:
What happens to culture when a dominant mode of knowledge has no temporal structure?
And the challenge ahead may not be understanding LLMs, but understanding how to restore tempo—how to recover the felt sense of time within a knowledge landscape that increasingly has none.
Restoring Tempo: Learning from Other Flat-Time Regimes
If LLMs inhabit an achronotopic regime—knowledge without duration—then one way to restore tempo is not by forcing them into human narrative time, but by looking to other flat-time knowledge systems that nevertheless developed an internal rhythm. The paradigmatic case is common law.
Common law is, at first glance, temporally strange. Judicial decisions remain perpetually present. A ruling from is not “old”; it is simply another voice in the present conversation. Like LLM outputs, common law rulings are timeless in the sense that they do not age—they accumulate. And yet common law does have tempo. It is not achronotopic. It has a very specific internal mechanism for generating temporal texture:
Landmark rulings become gravitational centers around which entire constellations of doctrine develop.
These centerpoints provide a rhythm of decision-making, an internal sense of priority and sequence, even though the domain itself does not unfold in ordinary historical time.
How Common Law Induces Tempo Without Sequence
Consider the doctrinal universe around the First Amendment in the United States.
Cases like:
Schenck v. United States (1919): introduced “clear and present danger”
Brandenburg v. Ohio (1969): replaced it with “imminent lawless action”
New York Times v. Sullivan (1964): established “actual malice” for public figures
Tinker v. Des Moines (1969): extended First Amendment protections to students
Citizens United v. FEC (2010): redefined political spending as protected speech
These rulings are not temporally sequential in the sense that courts “remember” them. Rather:
They create doctrinal nodes of heightened relevance.
Other cases are interpreted through them.
They accumulate jurisprudential weight.
They shape the decision-making tempo of subsequent courts.
The court does not say, “That case is from 1964, therefore it is old.” It says, “That case is controlling.” This is tempo without chronology. This is time induced through priority, not sequence. We might call this hierarchized flat time.
Priority as Tempo Management
In a regime where everything is present at once:
Priority becomes a proxy for temporality.
Hierarchy of relevance becomes a proxy for rhythm.
Weight becomes a proxy for duration.
A landmark case is not earlier than another case; it is heavier, more gravitational, more central. It structures what follows not because it preceded those cases, but because it frames them.
Tempo emerges through:
Canonicalizing certain rulings
Letting doctrine aggregate around them
Allowing exceptions to appear slowly as divergences
The court effectively creates a non-chronological temporality by elevating some rulings above others. This is exactly the kind of mechanism achronotopic systems require.
Toward Landmark Memory in LLMs
LLMs today operate across three distinct temporalities. First is the training freeze, a bulk and undifferentiated time in which vast historical archives are compressed into a single latent space. Second is the context window, an ephemeral, recency-biased present that disappears as soon as the conversation ends. And third is the emerging temporality created by retrieval-augmented generation (RAG) and in-inference learning, where the model pulls external documents or adapts behavior on the fly. This third temporality functions like a rapid micro-history: information is fetched, integrated, and acted upon in real time, but without ever forming a durable memory. Together these three temporalities produce a strange composite: a frozen past, a vanishing present, and a provisional just-in-time horizon that never accumulates into experience.
LLMs do now develop forms of memory—user-specific, task-specific, or system-level—but this memory is still a kind of unshaped flat time. It accumulates without hierarchy, without doctrinal weighting, without the equivalent of landmark decisions or canonical nodes. Nothing in the system yet distinguishes what should become “controlling precedent” from what should fade into the background.
If we treat LLM knowledge as akin to common law—polyphonic, precedent-based, inherently timeless—then restoring tempo requires the equivalent of landmark rulings inside the model.
This would mean:
Some memories must have higher internal priority than others.
Certain pieces of information must anchor decision trajectories.
Context must accumulate around stable doctrinal nodes.
The model must learn to treat some knowledge as gravitationally heavier.
This is the shift from: LLM memory as a 2-D archival space to LLM memory as a hierarchical jurisprudence
What matters is not chronology but organizational primacy. Just as “Brandenburg” controls modern speech doctrine while older cases fall away, an LLM might be guided to return to a stable, curated set of internal touchpoints when reasoning about: a domain, user, long-running projects, or a narrative thread.
Restoring Tempo in Achronotopic Thought
Achronotopic knowledge systems—common law, religion, LLMs—demonstrate that chronology is not the only way to experience time. Time can be simulated through:
priority
weight
interpretive pressure
the gravitational pull of canon
This is reassuring, because it suggests that tempo can be restored within LLM-mediated thought without forcing models into artificial narrative arcs.
What we need is an internal ecology of significance: a way for some ideas to be foregrounded, for others to be backgrounded, for certain nodes to collect interpretive mass.
In other words:
Tempo can be restored by giving the model its own “landmarks” — nodes of memory that acquire jurisprudential weight.
This transforms achronotopic knowledge into structured flat time, much like common law.
And ultimately, this redesign would allow LLMs to move from pure instantaneous polyphony toward a more stable, interpretable, and human-compatible temporal logic.
Brillaint framing of the temporal problem with LLMs. The common law analogy is espescially sharp because it shows that flat-time systems can stil have rhythm if they develop gravitational centers. The idea that landmark rulings create tempo through priority rather than sequence gets at something fundamental about how we organize knowledge when chronology isnt available. Makes me wonder if the shift from sequence to weight might actually be closer to how human memory works anyway.