Part IV: Remembering the Human Microcosm in the Age of Mechanized Intelligence

Part IV: Hegel's Loom and the Difference Reason Makes

Part I: The Pope Interrupts the Talking Machine

Remembering the Human Microcosm in the Age of Mechanized Intelligence

Part II: Philosophy as Emergency Response &
Part III: Resisting Cognitive Enclosure

Parts II & III: Remembering the Human Microcosm in the Age of Mechanized Intelligence

---

Part IV: Hegel’s Loom and the Difference Reason Makes

Metaphor is not just a shiny paint job on the vehicle of cognition. It is the engine of thought. Its coupling of concepts drives the limits of conceivability, shaping what is thought together and what is not thought at all. The metaphorical imagination is our main means of tuning in to the otherwise invisible effects of new media technologies. Part of the discipline philosophy brings is allowing us to notice an analogy as an analogy before advertising crystalizes it into the unnoticed transparency of common sense. A fact is a fact, but it might also be a fossilized metaphor. The governing analogy of our age is that cognition is computation: the brain an information-processing device, perception its input and behavior its output, memory a form of physical storage, learning the adjustment of weights, and intelligence an algorithm for minimizing error or surprisal. On this view, given enough training data and computational power, consciousness itself will eventually be engineered. It is a fertile analogy, that is not in dispute. It has generated new sciences and built the very technology now in question. But an analogy, strictly speaking, is not an identity but a proportional comparison: x is to y as z is to w. Metaphors often condense such analogies into assertions of identity. The metaphor “the mind is a computer,” for example, tacitly proposes that mind is to brain as software is to hardware. It may also imply that perception is to cognition as input is to processing, or that memory is to experience as data storage is to information, and so on. The metaphor gains its rhetorical force by suppressing the “as if” and presenting a selective structural comparison as though it were an identity. Reiterated in textbooks and earnings calls, in grant applications and policy briefs, the partial comparison congeals into an ontology, until we find ourselves insisting not that the mind is like a computer in some respects but that it simply is one—and, by the same illogic, that a sufficiently capable computer simply is a mind. Remembering the analogy and holding open where and whether it fits without pinching is the precondition of thinking clearly about our predicament, and of seeing the enclosure for what it is. We cannot feel robbed of a mind that we have been persuaded was only a machine all along.

Hegel, writing in the first third of the 19th century, offered a pre-cybernetic critique of the mind-machine metaphor by way of the anti-analogy of the loom. He thereby aids us in avoiding the misplaced concreteness that has by now all but hijacked our everyday thought. LLMs are like mechanical looms, not incarnations of Logos.

Some readers may be unfamiliar with Hegel, so is worth pausing to register how far his account of Reason departs from that of the better known Descartes, especially considering how much the computational metaphor owes to the latter. For Descartes, Reason is a faculty possessed by a thinking substance, the res cogitans, a solitary subject standing over against an extended world of mechanisms, certifying its representations one by one. Despite the potent intuitions and rich phenomenology of Descartes’ Meditations on First Philosophy (1641), once its methodology has been established, Cartesian rationality is ahistorical and proprietary, a reckoning device to be correctly operated rather than a living process to be undergone. Having evacuated mind from nature and confined it to the human head, Descartes turned the rest of the world, our own animal bodies included, into automata. The computational theory of mind later completed the Cartesian project by annexing the last remaining territory: if everything outside thought is mechanism, why not thought itself? Hegel’s work runs in precisely the opposite direction. Reason for him is not a static faculty housed in a worldless subject but a self-developing activity that includes its own history, an achievement of mediation formed through its embodied, linguistic, and social relations with a world that is never merely external to it. For Hegel, knowledge of truth cannot begin as the indubitable private property of an ego but is a result achieved only through the dialectical labor of negation, the ego’s transformative encounter with otherness. The Cartesian subject confronts its objects across an unbridgeable gap that only the assurance of God’s omnipotent goodness allows us to close. As Whitehead quipped, Descartes’ representative theory of perception forces him to resort to a deus ex machina, drawing on God’s power in “the crude form of giving a limited letter of credit” to establish the correctness of the ego’s representations.^[1] Hegelian Reason discovers that the gap is its own self-differentiation, a wound thought inflicts and heals in the same organic movement. Hegel thus radicalizes rather than simply repudiates Descartes: thought can know reality not because an isolated mind constructs an accurate internal picture of an alien world, but because reality itself is intelligibly articulated and comes to reflective consciousness in and through us. Descartes sought a method to render thought error proof, while Hegel understood that error and contradiction are internal to and required for thinking’s growth beyond its former limits. Reason that cannot fail, suffer its own mistakes, and be changed by what it thinks, is no Reason at all. It is, at best, a reliable loom.

Hegel followed Kant in distinguishing Reason (Vernunft) from the Understanding (Verstand). The former organically generates ideas in a process of dialectical development, while the latter can only mechanically rearrange ready-made concepts. Reason is not a calculator manipulating data in a fixed state space according to formal rules. Reason is self-moving and auto-generating. It directs and differentiates itself, encounters its own limits, negates its one-sidedness, and returns to itself transformed. The easy divisions between subject and object, form and content, syntax and semantics, and what we now call hardware and software are overcome in Hegel’s account of human thinking activity. Reason does not merely produce statistically ordered outputs. It—we, as rational animals—undergo the transformations we think. What this means is that, as Whitehead put it, “No thinker thinks twice; and, to put the matter more generally, no subject experiences twice.”^[2] This is unlike LLMs, which after their initial training runs and reinforcement tuning, must have their numerical weights frozen or risk catastrophic collapse.

In a remark in his Science of Logic (1813), specifically in the doctrine of Essence, Hegel articulates a critique of the isolated Understanding’s external mode of reflection by way of analogy to a mechanical loom. This mode of thought takes up the concepts of identity and difference ready-made, as though each was an already spun thread, and weaves them together. To imagine cognition as “nothing more than a loom,” weaving the warp of identity and the woof of difference, is to reduce thinking to a mechanism that works on already finished materials.^[3] The thinking subject and the objects thought are left fundamentally unchanged, no matter how intricate the pattern formed out of the threads. The loom’s textile products can be astonishingly complicated. But the loom itself stands at a remove from what it weaves. It combines given materials without inwardly transforming.

Hegel’s loom is the perfect analogy for large language models. LLMs are transformer-based machines for weighting relations among numerical patterns abstracted from symbols whose meaning belongs to human forms of life, and for generating contextually probable continuations of those patterns. They tokenize language, layer those tokens in complex mathematical vectors, and compute outputs by statistically reweaving relations learned from vast collections of text. An LLM’s architecture and capabilities are astonishing. It can generate discourse that bears a surface resemblance to judgment, poetry, even speculative philosophizing. But none of this amounts to cognition in the sense Hegel attributes to Reason. It is the production of probable outputs out of statistical patterns extracted from previously deposited meanings. It is the woven product of a loom, a textile whose meaningful threads precede and exceed it, having first been spun and later interpreted by human souls. Whether this difference leaves any measurable trace in the sense that the self-movement of Reason can be detected in the empirical comparison of how humans and LLMs read and write is a question I defer to a later section.

Crucially, for Hegel the Understanding is not simply the enemy of Reason but a partial moment within its self-development, a mode of thought that Reason takes up, negates in its one-sidedness, and preserves at a higher level. So the danger is not the Understanding as such, but the Understanding falsely absolutized, such that the part mistakes itself for the whole. Because we, too, spend much of our waking lives in the loom-like mode of the Understanding, we are especially prone to mistake mind for machine, and machine for mind. The more we offload our reasoning onto a tireless mechanical weaver, the more we are trained to think as it computes, until we are the more easily convinced that it computes as we thinks. This is why the remedy I propose is not to break into the mills under the cover of night to smash the looms, as the Luddites once did. The remedy is not to smash the loom but to sublate it, assuring that our cybernetic prostheses remain in service to human life, including the life of Reason.

A machine model of language can store, re-arrange, and transmit meaningful information. But it cannot create information or comprehend meaning. LLMs relay and recombine the fossilized and numerically tokenized traces of meaning. They take already expressed meanings, digitize and reweave them. They can do this with astonishing range and finesse. But the astonishingness of the relay of information should not be mistaken for its conscious creation. As we will learn from Ruyer in a later section, communication is never just the transmission of a pattern but involves an expressive and interpretive participation in meaning. Listening or reading is just as much a creative act as speaking or writing. Machines can assist us in our creative processes. But they cannot themselves create. The current generation of LLM tools should therefore be understood not as an independently emerging super-intelligence, but as the latest externalization and augmentation of human intelligence, one more chapter in our long co-evolutionary history with technology.

---

^[1] Whitehead, Process and Reality, 49.

^[2] Whitehead, Process and Reality, 29.

^[3] G. W. F. Hegel, The Science of Logic, trans. and ed. George di Giovanni (Cambridge: Cambridge University Press, 2010), 357 (GW 11:261).