||Roger Penrose's book Shadows of the Mind may be purchased
Can Physics Provide a Theory of Consciousness?
A Review of Shadows of the Mind by Roger Penrose
Bernard J. Baars
The Wright Institute
2728 Durant Ave.
Berkeley, Calif. 94704
Copyright (c) Bernard J. Baars 1995
PSYCHE, 2(8), May 1995
KEYWORDS: awareness; cognitive theory; consciousness; Global Workspace theory;
phenomenology; Quantum consciousness
REVIEW OF: Roger Penrose (1994) Shadows of the Mind. New York: Oxford
University Press. 457 pp. Price: $25 hbk. ISBN 0-19-853978-9.
1.1 Physics is surely the most beautiful of the sciences, and it is esthetically
tempting to suppose that two of the great scientific mysteries we confront
today, observer effects in quantum mechanics and conscious experience, are
in fact the same. Roger Penrose is an admirable contributor to modern physics
and mathematics, and his new book, Shadows of the Mind (SOTM) offers
us some brilliant intellectual fireworks --- which for me at least, faded
rapidly on further examination.
1.2 I felt disappointed for several reasons, but one obvious one: Is consciousness
really a physics problem? Penrose writes,
A scientific world-view which does not profoundly come to terms
with the problem of conscious minds can have no serious pretensions of completeness.
Consciousness is part of our universe, so any physical theory which
makes no proper place for it falls fundamentally short of providing a genuine
description of the world. I would maintain that there is yet no physical,
biological, or computational theory that comes very close to explaining
our consciousness ... (emphasis added)
1.3 Having spent 17 years of my life trying to do precisely what Penrose
suggests has not and cannot be done, this point was a bit disconcerting.
But even more surprising was the claim that consciousness is a problem in
physics. The conscious beings we see around us are the products of billions
of years of biological evolution. We interact with them --- with each other
--- at a level that is best described as psychological. All of our evidence
regarding consciousness depends upon reports of personal experiences, and
observation of our own perception, memories, attention, imagery, and the
like. The evidence therefore would seem to be exclusively psychobiological.
We will come back to this question.
1.4 The argument in SOTM comes down to two theses and a statement of faith.
The first thesis I will call the "Turing Impossibility Proof,"
and the second, the "Quantum Promissory Note". The statement of
faith involves classical Platonism of the mathematical variety, founded
in a sense of certainty and wonder at the amazing success of mathematical
thought over the last 25 centuries, and the extraordinary ability of mathematical
formalisms to yield deep insight into scientific questions (SOTM, p. 413).
This view may be captured by Einstein's well-known saying that "the
miraculous thing about the universe is that it is comprehensible."
While I share Penrose's admiration for mathematics, I do not believe in
the absolute nature of mathematical thought, which leads him to postulate
a realm of special conscious insight requiring no empirical investigation
to be understood.
1.5 After considering the argument of SOTM I will briefly sketch the current
scientific alternative, the emerging psychobiology of consciousness (see
Baars, 1988, 1994; Edelman, 1989; Newman and Baars, 1993; Schacter, 1990;
Gazzaniga, 1994). Though the large body of current evidence can be stated
in purely objective terms, I will strive to demonstrate the phenomena by
appealling to the reader's personal experience, such as your consciousness
of the words on this page, the inner speech that often goes with
the act of reading carefully, and so on. Such demonstrations help
to establish the fact that we are indeed talking about consciousness as
2. Has Science Failed To Understand Consciousness?
2.1 Central to SOTM is Penrose's contention that contemporary science has
failed to understand consciousness. There is more than a little truth to
that --- if we exclude the last decade --- but it is based on a great historical
misunderstanding: It assumes that psychologists and biologists have tried
to understand human experience with anything like the persistence and talent
routinely devoted to memory, language, and perception. The plain fact is
that we have not treated the issue seriously until very recently. It may
be difficult for physicists to understand this --- current physics does
not seem to be intimidated by anything --- but the subject of conscious
experience, the great core question of traditional philosophy, has simply
been taboo in psychology and biology for most of this century. I agree with
John Searle that this is a scandalous fact, which should be a great source
of embarassment to us in cognitive psychology and neuroscience. But no one
familiar with the field could doubt it. As Crick and Koch (1992) have written,
"For many years after James penned The Principles of Psychology
(1890) . . most cognitive scientists ignored consciousness, as did almost
all neuroscientists. The problem was felt to be either purely "philosophical"
or too elusive to study experimentally. . . In our opinion, such timidity
2.2 Fortunately the era of avoidance is visibly fading. First-order theories
are now available, and have not by any means been disproved (Baars,1983,
1988, and in press; Crick & Koch, 1992; Edelman, 1989; Gazzaniga, 1994;
Schacter, 1990; Kinsbourne, 1993; etc.). In fact, there are significant
commonalities among contemporary theories of consciousness, so that one
could imagine a single, integrative hybrid theory with relative ease. But
Penrose does not deal with this literature at all.
2.3 Has science failed, and do we need a scientific revolution? Given the
fact that we have barely begun to apply normal science to the topic, Penrose's
call for a scientific revolution seems premature at best. There is yet nothing
to revolt against. Of course we should be ready to challenge our current
assumptions. But it has not been established by any means that ordinary
garden-variety conscious experience cannot be explained through a
diligent pursuit of normal science.
3. A Critique Of The Turing "Impossibility Proof"
3.1 Impossibility arguments have a mixed record in science. On one side
is the proof scribbled on the back of an envelope by physicists in the Manhattan
Project, showing that the first enriched uranium explosion would not trigger
a chain reaction destroying the planet. But notice that this was not a purely
mathematical proof; it was a physical-chemical-mathematical reductio, with
a very well-established, indispensible empirical basis. On the side of pure
mathematics, we have such historical examples as Bishop Berkeley's disproof
of Newton's use of infinitesimals in the calculus. Berkeley was mathematically
right but the point was empirically irrelevant; physicists used the flawed
calculus for two hundred years with great scientific success, until just
before 1900 the paradox was resolved by the discovery of converging series.
3.2 Even more empirically irrelevant was Zeno's famous Paradox, which seemed
to show that we cannot walk a whole step, since we must first cover half
a step, then half of half a step, then half of the remaining distance, and
the like, never reaching the whole intended step. Zeno of Elea used this
clever argument to prove to the astonishment of the world that motion was
impossible. But that did not paralyze commerce. Ships sailed, people walked,
and camels trudged calmly on their way doing the formally impossible thing
for a couple of thousand years until the formal solution emerged. And of
course we have more than a century of mathematical reductios claiming that
Darwinian evolution is impossible if you combine all the a priori probabilities
of carbon chains evolving into DNA and ending up with thee and me. These
reductios on behalf of divine Creation still appear with regularity, but
the biological evidence is so strong that they are not even considered.
3.3 The problem is of course that a mathematical model is only as good as
its assumptions, and those depend upon the quality of the evidence. The
whole Turing Machine debate and its putative implications for consciousness
is in my opinion a great distraction from the sober scientific job of gathering
evidence and developing theory about the psychobiology of consciousness
(e.g., Baars, 1988; 1994). The notion that the Turing argument actually
tells us something scientifically useful is amazingly vulnerable. After
all, the theory assumes an abstract automaton blessed with infinite time,
infinite memory, and an environment that imposes no resource constraints.
The brain is a massively parallel organ with 100 billion simultaneously
active neurons, but the Turing Machine is at the extreme end of serial machines.
This appears to be the reason why discussion of the Turing topic appears
nowhere in the psychobiological literature. It seems primarily limited to
philosophy and the general intellectual media.
3.4 Finally, it turns out that all current cognitive and neural models are
formal Turing equivalents. That means the mathematical theory is useless
in the critical task of choosing between models that are quite different
computationally and on the evidence. It does not distinguish between neural
nets and symbolic architectures for example, as radically different as they
are in practice. But that is exactly the challenge we face today: choosing
between theories based on their fit with the evidence. Here the theory of
automata is no help at all.
3.5 A small but telling fact about Penrose's book caught my attention: of
its more than 400 references, fewer than forty address the psychology or
biology of consciousness. But all our evidence on the subject is psychological
and, to a lesser extent, biological! It appears that Penrose's topic is
not consciousness in the ordinary psychoneural sense, like waking up in
the morning from a deep sleep or listening to music. How the positive proposals
in SOTM relate to normal psychobiological consciousness is only addressed
in terms of a technical hypothesis. Stuart Hameroff, an anesthesiologist
at the University of Arizona currently working with Penrose, has proposed
that general anesthetics interact with neurons via quantum level events
in neural microtubules, which transport chemicals down axons and dendrites.
It is an interesting idea, but it is by no means accepted, and there are
many alternative hypotheses about anesthetics. But it is a real hypothesis:
testable, relevant to the issue of consciousness, and directly aimed at
the quantum level.
3.6 Penrose calls attention to the inability of Turing Machines to know
when to stop a possibly nonterminating computation. This is a form of the
Goedel Theorem, from which Penrose draws the following conclusion: "Human
mathematicians are not using a knowably sound algorithm in order to ascertain
mathematical truth." That is to say, if humans can propose a Halting
Rule which turns out to be demonstrably correct, and if we take Turing Machines
as models of mathematicians, then the ability of mathematicians to come
up with Halting Rules shows that their mental processes are not Turing-computable.
3.7 I'm troubled by this argument, because all of the cognitive studies
I know of human formal reasoning and logic show that humans will take any
shortcut available to find a plausible answer for a formal problem; actually
following out formalisms mentally is rare in practice, even among scientists
and engineers. Human beings are not algorithmic creatures; they prefer by
far to use heuristic, fly-by-the-seat-of-your-pants analogies to situations
they know well. Even experts typically use heuristic shortcuts. Furthermore,
the apparent reductio of Penrose's claim has a straightforward alternative
explanation, namely that one of the premises is plain wrong. The implication
psychologically is not that people are fancier than any Turing Machine,
but that they are much sloppier that any explicit algorithm, and yet do
quite well in many cases.
3.8 The fact that people can walk is an effective counter to Zeno's
Paradox. The fact that people can talk in sentences was Chomsky's counter
to stimulus-response theories of language. Now we know that people can in
many cases find Halting Rules. It's not that human processes are noncomputible
by a real computer --- numerous mental processes have been simulated with
computers, including some formidable ones like playing competitive chess
--- but rather that the formal straightjacket of Turing Machinery is simply
the wrong model to apply. This is the fallacy in trying to attribute rigorous
all-or-none logical reasoning to ordinary human beings, who are pragmatic,
heuristic, cost-benefit gamblers when it comes to solving formal problems.
3.9 Penrose proceeds to deduce that consciousness is noncomputable by Turing
standards. But even this claim is based only on intuition; the argument
has the form, "mathematicians have an astonishingly good record gaining
fundamental insights into a variety of formal systems; this is obviously
impossible for a Turing automaton; hence mathematicians themselves cannot
be modeled by such automatons." From a psychobiological point of view
the success of mathematical intuition is more likely reflect the nervous
system's excellent heuristics for discovering patterns in the world. The
brain appears to have sophisticated knowledge of space, for example, which
may in turn allow deep geometrical intuitions to occur with great accuracy
in talented individuals. In effect, we may put a billion years of brain
evolution of spatial processing to good use if we are fortunate enough to
be mathematically talented.
4. The Quantum Promissory Note
4.1 Having proved that Turing machines cannot account for mathematical intuition,
Penrose develops the idea that Quantum Mechanics will provide a solution.
QM is the crown jewel of modern theoretical physics, an endless source of
insight and speculation. It shows extraordinary observer paradoxes. Consciousness
is a mysterious something human observers have, and many people leap to
the inference that the two observer mysteries must be the same. But this
is at best a leap of faith. It is much too facile: observations of quantum
events are not made directly by human beings but by such devices as Geiger
counters with no consciousness in any reasonable sense of the word. Conscious
experience so far as we know is limited to huge biological nervous systems,
produced over a billion years of evolution.
4.2 There is no precedent for physicists deriving from QM any macrolevel
phenomenon such as a chair or a flower or a wad of chewing gum, much less
a nervous system with 100 billion neurons. Why then should we believe that
one can derive psychobiological consciousness from QM? QM has not been shown
to give any psychological answers. Conscious experience as we know it in
humans has no resemblance to recording the collapse of a quantum wave packet.
Let's not confuse the mysteries of QM with the question of the reader's
perception of this printed phrase , or the inner sound of these
4.3 What can we make of Penrose's Quantum Promissory Note? All scientific
programs are promissory notes, making projections about the future and betting
on what we may possibly find. The Darwin program was a promissory note,
the Human Genome project is, as are particle physics and consciousness research.
How do you place your bets? Is there a track record? Is there any evidence?
5. Treating Consciousness As A Variable: The Evidence For Consciousness
5.1 We are barely at the point of agreeing on the real scientific questions,
and on the kind of theory that could address them. On the matter of evidence,
Baars (1983, 1988, 1994 and in press), Libet (1985) and others have argued
that empirical constraints bearing on consciousness involve a close comparison
of very similar conscious and unconscious processes. As elsewhere in science,
we can only study a phenomenon if we can treat it as a variable. Many scientific
breakthroughs result from the realization that some previously assumed constant,
like atmospheric pressure, frictionless movement,the uniformity of space,
the velocity and mass of the Newtonian universe, and the like, were actually
variables, and that is the aim here. In the case of consciousness we can
conduct a contrastive analysis comparing waking to sleep, coma, and general
anesthesia; subliminal to supraliminal perception, habituated vs. novel
stimuli, attended vs. nonattended streams of information, recalled vs. nonrecalled
memories, and the like. In all these cases there is evidence that the conscious
and unconscious events are comparable in many respects, so that we can validly
probe for the essential differences between otherwise similar conscious
and unconscious events (See Greenwald, 1992; Weiskrantz, 1986; Schacter,
5.2 This "method of contrastive analysis" is much like the experimental
method: We can examine closely comparable cases that differ only in respect
to consciousness, so that consciousness becomes, in effect, a variable.
However, instead of dealing with only one experimental data set, contrastive
analysis involves entire categories of well-established phenomena, summarizing
numerous experimental studies. In this way we can highlight the variables
that constrain consciousness over a very wide range of cases. The resulting
robust pattern of evidence places major constraints on theory (Baars, 1988;
6. Can Penrose Deal With Unconscious Information Processing?
6.1 Like many psychologists before 1900 Penrose appears to deny unconscious
mental processes altogether. This is apparently because his real criterion
is introspective access to the world of formal ideas. But introspection
is impossible for unconscious events, and so the tendency for those who
rely on introspection alone is to disbelieve the vast domain of unconscious
6.2 Unconscious processing can be inferred from numerous sources of objective
evidence. The simplest case is the great multitude of your memories that
are currently unconscious. You can now recall this morning's breakfast ---
but what happened to that memory before you brought it to mind? There is
much evidence that even before recall the memory of breakfast was still
represented in the nervous system, though not consciously. For example,
we know that unconscious memories can influence other processes without
ever coming to mind. If you had orange juice for breakfast today you may
switch to milk tomorrow, even without bringing today's juice to mind. A
compelling case can be made for unconscious representation of habituated
stimuli, of memories before and after recall, automatic skills, implicit
learning, the rules of syntax, unattended speech, presupposed knowledge,
preconscious input processing, and many other phenomena. In recent years
a growing body of neurophysiological evidence has provided convergent confirmation
of these claims. Researchers still argue about some of the particulars,
but it is widely agreed that given adequate evidence, unconscious processes
may be inferred.
6.3 What is the critical difference then between comparable conscious and
unconscious processes? There are several, but perhaps the most significant
one is that conscious percepts and images can trigger access to unanticipated
knowledge sources. It is as if the conscious event is broadcast to memory,
skill control, decision-making functions, anomaly detectors, and the like,
allowing us to match the input with related memories, use it as a cue for
a skilled actions or decisions, and detect problems in the input. At a broad
architectural level, conscious representations seem to provide access to
multiple knowledge source in the nervous system, while unconscious ones
seem to be relatively isolated. The same conclusion follows from other contrastive
analyses. (See Baars, 1988).
6.4 None of this evidence appears to fit in the SOTM framework, because
it has no role for unconscious but vitally important information processing.
This is a major point on which the great weight of psychobiological evidence
and SOTM are fundamentally at odds.
7. The Emerging Psychobiology Of Consciousness
7.1 The really daring idea in contemporary science is that consciousness
may be understandable without miracles, just as Darwin's revolutionary
idea was that biological variation could be understood as a purely natural
phenomenon. We are beginning to see human conscious experience as a major
biological adaptation, with multiple functions. It seems as if a conscious
event becomes available throughout the brain to the neural mechanisms of
memory, skill control, decision-makings, anomaly detection, and the like,
allowing us to match our experiences with related memories, use them as
a cue for skilled actions or decisions, and detect anomalies in them. By
comparison, unconscious events seem to be relatively isolated. Thus consciousness
is not just any kind of knowledge: It is knowledge that is widely distributed,
that triggers off widespread unconscious processing, has multiple integrative
and coordinating functions, aids in decision-making, problem-solving and
action control, and provides information to a self-system.
8.1 I don't know if consciousness has some profound metaphysical relation
to physics. Science is notoriously unpredictable over the long term, and
there are tricky mind-body paradoxes that may ultimately demand a radical
solution. But at this point in the vexed history of the problem there is
little question about the preferable scientific approach. It is not to try
to solve the mind-body problem first --- that effort has a poor track record
--- or to pursue lovely but implausible speculations. It is simply to do
good science using consciousness as a variable, and investigating its relations
to other psychobiological variables.
Baars, B.J. (1983). Conscious contents provide the nervous system with coherent,
global information. In R. Davidson, G. Schwartz, &
D. Shapiro (Eds.), Consciousness and self-regulation, 3, 45-76. New
York: Plenum Press.
Baars, B.J. (1988) A cognitive theory of consciousness. Cambridge,
UK: Cambridge University Press.
Baars, B.J. (1994) A thoroughly empirical approach to consciousness. PSYCHE
1(6) [80 paragraphs] URL:http://psyche.cs.monash.edu.au/volume1/ psyche-94-1-6-contrastive-1-baars.html
Baars, B.J. (in press) Consciousness regained: The new science of human
experience. Oxford, UK: Oxford University Press.
Crick, F.H.C. & Koch, C. (1992) The problem of consciousness, Scientific
American, 267(3), 153-159.
Edelman, G. (1989) The remembered present: A biological theory of consciousness.
NY: Basic Books.
Gazzaniga, M. (1994) Cognitive neuroscience. Cambridge, MA: MIT Press.
Greenwald, A. (1992). New Look 3, Unconscious cognition reclaimed. American
Psychologist, 47(6), 766-779.
James, W. (1890/1983). The principles of psychology. Cambridge, MA:
Harvard University Press.
Kinsbourne, M. (1993). Integrated field model of conscousness. In G. Marsh
& M. J. Brock (Eds.), CIBA symposium on experimental and theoretical
studies of consciousness. (pp. 51-60). London: Wiley Interscience.
Libet, B. (1985) Unconscious cerebral initiative and the role of conscious
will in voluntary action. Behavioral and Brain Sciences, 8, 529-66.
Newman, J., & Baars, B. J. (1993). A neural attentional model for access
to consciousness: A Global Workspace perspective. Concepts In Neuroscience,
Penrose, R. (1994) Shadows of the mind. Oxford, UK: Oxford University
Schacter, D. L. (1990). Toward a cognitive neuropsychology of awareness:
Implicit knowledge and anosognosia. Journal of Clinical and Experimental
Neuropsychology, 12(1), 155-178.
Weiskrantz, (1986) Blindsight: A single case and its implications.
Oxford, UK: Clarendon Press.