Overview and place in Penrose’s work
Shadows of the Mind: A Search for the Missing Science of Consciousness is a book by the British mathematical physicist Roger Penrose, published by Oxford University Press in 1994. It is a work of scientific and philosophical argument aimed at a serious general readership, though parts of it are technically demanding and engage with mathematical logic, the foundations of quantum mechanics, and neuroscience. The book functions as a sequel to and defense of Penrose’s earlier and more famous The Emperor’s New Mind (1989), which had made the controversial claim that human consciousness cannot be a purely computational or algorithmic process. Shadows of the Mind returns to that thesis, restates it in what Penrose regarded as a sharper and more rigorous form, responds at length to the many objections raised against the earlier book, and develops in greater detail the positive proposal that a genuinely new physics—rooted in quantum mechanics and its relationship to gravity—will be needed to account for the mind.
Penrose came to these questions with an unusual scientific pedigree. He is a mathematician and mathematical physicist known for foundational contributions across several fields: work on gravitational collapse and spacetime singularities (partly in collaboration with Stephen Hawking), the invention of nonperiodic tilings that bear his name, twistor theory, and diagrammatic techniques in relativity. Decades after Shadows of the Mind, he shared the 2020 Nobel Prize in Physics for his theoretical work establishing that black hole formation is a robust prediction of general relativity. This background matters for understanding the book, because Penrose approaches the problem of consciousness not as a philosopher of mind or a cognitive scientist by training, but as a physicist convinced that the deepest puzzles of mind are ultimately continuous with the unsolved puzzles of fundamental physics—especially the unresolved problem of how quantum mechanics is to be reconciled with general relativity and how the “collapse” of the quantum wavefunction actually occurs.
The central thesis: understanding is non-computational
The core claim of Shadows of the Mind is that human mathematical understanding is not, and cannot be, the product of any algorithm—any sound, knowably sound, computational procedure. From this Penrose argues, more broadly, that conscious understanding as such is a non-computational activity, and therefore that the project of “strong artificial intelligence,” which holds that the mind is essentially a computer program that could in principle be implemented on any suitable machine, is mistaken in principle rather than merely in practice.
To organize the debate, Penrose lays out four possible positions on the relationship between consciousness and computation. These are the well-known “four viewpoints” of the book:
- Viewpoint A holds that all thinking is computation, and that the right kind of computation would itself evoke conscious awareness. This is the position of strong AI.
- Viewpoint B holds that awareness is a genuine feature of the physical activity of the brain, and although any such physical activity can be simulated computationally, the simulation itself would not produce awareness; the computation would merely describe, not instantiate, consciousness.
- Viewpoint C, which is Penrose’s own, holds that the relevant physical activity of the brain that produces awareness cannot even be properly simulated by computation at all. On this view consciousness is a physical, scientifically explicable phenomenon, but the physics involved is non-computable.
- Viewpoint D holds that awareness cannot be explained in physical, computational, or any scientific terms whatsoever—a mystical or dualist position.
Penrose positions himself firmly at C: he is neither a strong-AI computationalist nor a mysterian. He insists that consciousness is a natural, physical phenomenon that science can and must explain, but that the explanation will require physics that lies beyond the reach of computation as we currently understand it.
The Gödelian argument
The intellectual centerpiece of the first part of the book is an argument drawn from mathematical logic, and specifically from Kurt Gödel’s incompleteness theorems. This line of reasoning had earlier been advanced in a different form by the philosopher J. R. Lucas, and Penrose develops his own more elaborate version.
The argument, in outline, is that for any sufficiently powerful and consistent formal system of the kind that could in principle capture the workings of a mathematician’s reasoning as an algorithm, Gödel’s methods allow one to construct a statement that is true but not provable within that system. Penrose contends that a human mathematician can, by reflecting on the system and grasping its meaning, “see” or understand that such a Gödel sentence is true, even though the algorithm itself cannot establish it. If human mathematical insight could always outstrip any given formal system in this way, then—Penrose argues—that insight cannot be identical to the operation of any single, fixed, knowably sound algorithm. He draws the conclusion that the faculty of mathematical understanding is non-algorithmic.
Penrose is careful, in Shadows of the Mind, to try to close what he saw as loopholes in the cruder versions of the argument. In particular he engages with the objection that a human mathematician might be an unsound or inconsistent algorithm, or an algorithm whose specification is unknowable even to the mathematician who instantiates it. Much of the detailed reasoning of the first half of the book is devoted to arguing that these escape routes are not, in the end, tenable, and that the conclusion of non-computability survives careful scrutiny.
Toward a new physics of mind: quantum theory and the brain
Having argued in the first part that consciousness involves non-computational processes, Penrose turns in the second part of the book to the question of where in the physical world such non-computable behavior might originate. His answer draws on his long-standing conviction that the standard formalism of quantum mechanics is incomplete, particularly with respect to the “measurement problem”—the question of how and why a quantum system, described by a smoothly evolving and superposed wavefunction, comes to yield a single definite outcome when observed.
Penrose distinguishes between the deterministic, computable evolution of the quantum state (described by the Schrödinger equation, which he labels with the letter U for unitary evolution) and the apparently discontinuous “reduction” or collapse of the state (which he labels R). He argues that reduction is a real, objective physical process rather than a mere artifact of observation or a bookkeeping device, and that it is here that non-computable physics may enter. Crucially, Penrose ties this objective reduction to gravity: he proposes that when a quantum superposition involves a large enough difference in the distribution of mass—and hence a large enough difference in spacetime curvature between the superposed alternatives—the superposition becomes unstable and spontaneously reduces to a single state. On this view the missing ingredient is a theory of quantum gravity, and the point at which superpositions collapse is set by a threshold related to gravitational self-energy. Penrose suggests that the correct, as yet undiscovered, theory of this “objective reduction” would be non-computable, thereby supplying the physical basis for the non-computational character of understanding argued for in the first part.
The book then seeks a plausible location within the brain where such quantum processes could be relevant to cognition. Here Penrose draws on ideas developed in collaboration with the anesthesiologist Stuart Hameroff, focusing on the microtubules—protein structures forming part of the cytoskeleton within neurons. Penrose and Hameroff proposed that microtubules could support large-scale quantum coherent states, and that the orchestrated collapse of such states could be the physical substrate of conscious moments. This proposal later became known as “orchestrated objective reduction” (Orch-OR). In Shadows of the Mind the neurobiological aspect is presented as a tentative and speculative suggestion—an attempt to indicate that the required non-computational quantum physics might have somewhere biological to act—rather than as an established mechanism.
Reception, criticism, and significance
Shadows of the Mind was widely reviewed and generated substantial and often sharp debate across several disciplines. Because it advanced a strong and unfashionable thesis against the dominant computational view of mind, and because it did so by combining arguments from three difficult fields—mathematical logic, quantum foundations, and neuroscience—it drew criticism from specialists in each.
The Gödelian argument attracted the most concentrated response from logicians and philosophers. Critics such as Solomon Feferman and Hilary Putnam, among others, challenged the inference from Gödel’s theorems to the non-computability of human thought, questioning in particular Penrose’s handling of the assumptions about the soundness and knowability of the putative algorithm, and arguing that the logical results do not license the conclusions Penrose draws about the human mind. This revived a long-running controversy that had earlier surrounded Lucas’s version of the argument.
The physical and biological proposals drew criticism of a different kind. A frequently cited objection concerned whether quantum coherence could survive in the warm, wet, and noisy environment of the brain for anything like the timescales that would be needed for it to be functionally relevant; the physicist Max Tegmark, for instance, argued that decoherence would destroy the relevant quantum states far too quickly. Many neuroscientists were skeptical of assigning a special cognitive role to microtubules. Defenders and collaborators, including Hameroff, responded to these criticisms and continued to develop the Orch-OR proposal in subsequent years.
Despite—and partly because of—this controversy, Shadows of the Mind became one of the most discussed books in the debate over consciousness and artificial intelligence in the 1990s. Its lasting significance lies less in any consensus that it produced (it produced none) than in the seriousness and prominence with which a major physicist argued that the mind might require a genuinely new physics, and that the reigning computational orthodoxy might be fundamentally mistaken. The book articulated a distinctive third position between reductive computationalism and outright mysticism, insisting on both the physical, scientific explicability of consciousness and its non-algorithmic character. It stands, alongside The Emperor’s New Mind, as the fuller statement of Penrose’s views on mind, and it helped keep alive interlinked questions about Gödel’s theorems, the measurement problem in quantum mechanics, and the limits of computation as central topics in the interdisciplinary study of consciousness.
