Merlin Donald

In the [YOU] on AI Field Guide

The cycle that began with [YOU] on AI frames the AI transition through the metaphor of the river of intelligence: a flow that has been running for 13.8 billion years, widening with each new channel, now arriving at a moment when the channel itself becomes active rather than merely present. Donald's three-transition framework is the most rigorous available account of the river metaphor. He shows that every previous widening followed a consistent structural pattern: driven by a change in the external cognitive environment rather than the biological brain, adding a new layer without replacing previous layers, producing genuine gains alongside genuine losses, resisted by practitioners adapted to the previous configuration, and resolved only by institutional construction that took, in each case, generations to complete. The cycle's Stage Four—the adaptation stage, the building of dams—is Donald's account of what institutional response looks like, and his historical record of how long it has taken and how badly it has gone wrong is the most precise measure of the urgency of the present moment.

His concept of layer collapse—the progressive neglect of foundational cognitive layers when a powerful upper layer becomes available—names what the cycle elsewhere calls the atrophy of ascending friction. The bards who could hold fifteen thousand lines of the Iliad in memory became obsolete within generations of widespread literacy. The scribes' patient manual engagement with text retreated from the center of intellectual life to its decorative margins. The navigators who could plot a course by the stars are being replaced by GPS-dependent pilots who cannot. Each transition produced the same pathology: capacities not exercised by the new medium atrophied across the population, often invisibly, discovered only when a crisis required them and found them unavailable. Donald's gift to the present moment is the precise structural description of this pathology, which lets us identify it before it becomes irreversible rather than after.

The cycle's account of the imagination-to-artifact ratio—the collapse of the distance between conceiving something and making it—maps cleanly onto Donald's analysis of what changes when the external memory field becomes active. In every previous cognitive configuration, the human mind was the only active processor in a system of passive storage. Writing stored information; it did not generate it. A database organized information; it did not interpret it. The translation cost between internal thought and external implementation consumed most of the cognitive bandwidth of every knowledge worker in the theoretic era. When the external field becomes active—when the thought can become the thing without passing through layers of specification, communication, and manual execution—those cognitive resources are freed for judgment, direction, and taste. Donald's framework predicts this reorganization as the structural consequence of the transition, and it names the danger that follows: the capacities freed from translation may not be developed into judgment if the lower layers that built judgment—the mimetic engagement with materials, the mythic encounter with difficulty—are simultaneously atrophying.

Origin

Merlin Donald was born in 1939 and spent most of his career at Queen's University in Kingston, Ontario, where he taught psychology and cognitive science. His intellectual project emerged from a dissatisfaction with the standard accounts of what makes human cognition distinctive. The dominant answer in cognitive science—that humans are smarter because we have bigger or better brains—foundered on the simple observation that the biological brain has not changed in a hundred thousand years. Whatever explains the difference between the hunter and the physicist cannot be in the neurons. Donald's answer, developed across his two major works, is that the explanation is in the media through which cognition operates, and that these media have a history with a specific structure.

His three-stage theory distinguishes the mimetic revolution (the use of the body as a representational medium, dating to perhaps two million years ago), the mythic revolution (oral language and narrative, perhaps one hundred thousand years ago), and the theoretic revolution (external symbolic storage—writing, mathematics, notation—dating to about five thousand years ago). Each transition was driven not by genetic change but by the invention of a new external medium, and each produced a new kind of mind by changing the cognitive ecosystem within which the biologically unchanged brain operated. The physicist's brain is the hunter's brain operating within an ecosystem that includes libraries, mathematical notation, peer-reviewed literature, and computational tools. Remove the ecosystem and the cognitive distance collapses.

Donald's framework became influential in cognitive science, philosophy of mind, and educational theory, and it has been independently rediscovered in adjacent forms by researchers in distributed cognition, cultural psychology, and the philosophy of technology. His work anticipates, with unusual precision, the questions that artificial intelligence now forces on every field that thinks about human cognitive development—questions about what belongs inside the skull and what belongs in the environment, about what capacities require embodied practice and what can be transmitted through symbolic media, and about the institutional conditions under which cognitive transitions produce flourishing rather than fragmentation.

Key Ideas

The three cognitive revolutions. The mimetic revolution introduced the body as a representational medium—gesture, ritual, dance, manual skill, the intentional re-enactment of past events. The mythic revolution introduced oral language and narrative, externalizing thought into sound and organizing it into temporal sequences with causal structure. The theoretic revolution introduced external symbolic storage: writing, mathematics, diagrams, libraries, the entire apparatus that allows cognition to operate outside the skull, to be inspected, revised, accumulated, and transmitted without degradation. Each layer is cumulative: the modern mind carries all three simultaneously, and the richest cognitive experiences engage all three at once.

The external memory field. Donald's term for the culturally constructed environment of texts, tools, and symbolic systems that extends individual cognitive reach beyond what the biological brain alone can achieve. The crucial feature of every external memory field before AI was that it was passive—it stored information but did not generate it, organized knowledge but did not create it. The human brain was the only active processor at the center of a passive storage network. When the field becomes active, as it does with AI, the cognitive architecture changes at the most fundamental level. This is not a faster tool; it is a new kind of cognitive ecosystem, and the transition is as consequential as the original invention of writing.

Layer collapse. Every cognitive transition produced a characteristic pathology: the progressive neglect of the cognitive layer displaced by the new arrival, leading to the atrophy of capacities that are no longer exercised because they are no longer necessary. The invention of writing atrophied prodigious oral memory. The printing press marginalized the scribes' intimate knowledge of text through manual engagement. GPS is measurably eroding spatial navigation ability in populations that rely on it. Layer collapse is not dramatic; it is incremental, invisible, and driven by the accumulation of individually rational decisions. Its aggregate effect is visible only when a crisis requires the atrophied capacity and finds it unavailable.

The cognitive hybrid. Donald's framework provides the theoretical basis for understanding what happens when a human mind works in sustained collaboration with an active AI system. The result is not a human using a tool; it is a hybrid cognitive system with capabilities neither participant possesses alone. Every previous cognitive transition produced such a hybrid—the mind-plus-mimetic-culture system, the mind-plus-language system, the mind-plus-external-storage system. The mind-plus-AI system is the newest, and its distinctive danger is that sustained operation within the hybrid may degrade the human component's independent capacities in the ways that literacy degraded oral memory: gradually, invisibly, until the capacity is needed and found to be unavailable.

Institutional management as the variable. The long-term outcome of every cognitive transition depended not on the technology but on the institutional structures built to manage it. Writing could have been and often was confined to priestly elites. The printing press could have produced only propaganda. Donald's historical record is unambiguous: the transitions that were managed with adequate institutional investment in the lower layers—the educational practices, professional norms, and cultural expectations that required continued engagement with mimetic, mythic, and theoretic cognition even as the algorithmic layer became available—produced flourishing. The transitions that were not produced fragmentation and loss that took generations to repair. The present transition allows no such generations.