Penrose on Why Consciousness Can't Be a Computation

Penrose opens by stating his position plainly: whatever consciousness is, it is not a computation. It is a physical process, he says, but not one that can be described by computing. Fridman raises the orchestrated objective reduction proposal, and Penrose first replies that he had no idea how to get there when he started.

He had always planned to write a semi-popular book once he retired, explaining the puzzles of physics, mathematics and computability. A radio discussion between Edward Fredkin and Marvin Minsky pushed him to start sooner. They pictured a room where two computers talk to each other and, by the time you reach them, have exchanged more ideas than the whole human race ever has.

Penrose understood where they were coming from but felt something was missing, so he wrote *The Emperor's New Mind*. He notes that it appeared at about the same time as Stephen Hawking's *A Brief History of Time*, and that Martin Gardner supplied a very nice foreword for his book. Fridman observes that Fredkin and Minsky represent the AI camp's hopes and dreams, and Penrose agrees: he saw their perspective but did not share it.

While writing, Penrose realized he knew little about neurophysiology and started reading up. He wanted to know how nerve signals could preserve quantum coherence, but everything he read said that the electrical signals travelling along nerves would disrupt any such effect. Near the end of the book he had more or less given up.

He had hoped the book would inspire young people, but the letters he received came mostly from older readers, who had the time to read it. The exception was Stuart Hameroff, who told him he was missing something: microtubules. Penrose recalls thinking this was a much more promising structure.

Microtubules are the small tubes that pull chromosomes apart when a cell divides, which is what first caught Hameroff's interest. Hameroff's day job is as an anesthesiologist, whose task is to turn consciousness off and back on reversibly. He formed the view that anesthetic gases act on microtubules, though Penrose admits he is unclear on the details of why.

Microtubules are found in nearly all cells in the body, apart from red blood cells, but they come in different kinds. The ones that excited Penrose most were the only ones he knew of at the time. They are very symmetrical, and he had reason to believe such structures would be much better at preserving a quantum state.

The point is to keep certain degrees of freedom from leaking into the environment, since once they leak the quantum state is lost. Penrose wants these states preserved at a level where state reduction can occur, and that is where he locates the non-computability. The measurement process in quantum mechanics is the thing he thinks matters here.

Fridman then summarizes the idea for newcomers: orchestrated objective reduction, or Orch OR, holds that consciousness originates at the quantum level inside neurons. That contrasts with the view that it arises from computation performed by synapses. Penrose adds that the symmetry argument impressed him a great deal, though he did not know at the time about the difference between A and B lattices in microtubules.

Penrose stresses that his claim is more radical than saying biology uses quantum mechanics. All of chemistry is quantum mechanical, and photosynthesis appears to use quantum effects in a basic way, so quantum biology is no longer shocking. But his argument concerns conventional quantum mechanics, and following the Schrödinger equation is still computable. You have to go to the place where quantum mechanics goes wrong.

He explains that standard practice mixes two processes. One is the deterministic Schrödinger evolution, which leads to absurd situations such as Schrödinger's cat being dead and alive at once. Penrose says Schrödinger raised the cat as an absurdity, a sign that the reduction of the state, or collapse of the wavefunction, needs to be understood and does not follow the equation.

He points to Einstein, Schrödinger and Dirac as great physicists who regarded quantum mechanics as provisional and in need of something to explain collapse. Penrose belongs to a small group of mavericks who think the theory must be modified, and to an even smaller minority who think gravity is the key. Quantum gravity as usually pursued applies quantum mechanics to gravity, whereas he wants a two-way relationship in which gravity also affects the structure of quantum mechanics. He admits we have no such theory yet.

Asked whether any experiments exist, Penrose says some are being thought through and he hopes they will be performed. One is being developed by his long-time acquaintance Jeremy Stowe, who splits his time between Leiden and Santa Barbara. It aims to show that quantum mechanics as we understand it must be modified if gravitational effects are not brought in. Other experiments, he adds, look at the microtubule side in biology.

Penrose calls Hameroff's work an important area. One of the few ways to get a handle on consciousness is to look at what turns it off, and general anesthetics do this very specifically. Hameroff and his collaborators believe the gases directly affect microtubules, and Penrose says there is some evidence for it, though he cannot say how strong or watertight it is.

Fridman notes that few people are working on this, and Penrose agrees it is a minority view but says it is no longer zero. In the early days it was only him and a few friends. Now it is one of perhaps four, five or six views that people hold about consciousness.

Penrose recalls describing physics theories as superb, useful or tentative. Fridman asks whether we will get a superb theory of consciousness, and Penrose answers that we might, but we are a long way off and probably not even at the tentative stage. Most views, he says, are computational in some form, and even the integrated information theory camp talks of consciousness as not computation. His response is to ask what it then is, if physical processes are going on.

He contrasts his view with an older idea held by some distinguished quantum physicists, that a conscious observer looking at a system is what reduces the state. Penrose proposes almost the reverse: the state reduces itself in some non-computational way we lack a theory of, and that choice is a building block of consciousness. Each such moment is an element of proto-consciousness, which when organized together makes up actual consciousness.

The "OR" in Orch OR, the quantum choice, is something he can see how to develop as a theory. The "Orch" part, the orchestration of these events into real conscious experience, is much more mysterious. He compares the word to "material", a placeholder that hides how little we know. Finally he places understanding at the center of awareness, understanding and intelligence: he believes intelligence needs understanding and understanding needs awareness, and as a mathematician he concentrates on understanding because that is where he can say something.