(Perspectives in Cognitive Neuroscience) Benjamin Libet - Mind Time - The Temporal Factor in Consciousness (2005, Harvard University Press)

PERSPECTIVES IN COGNITIVE NEUR
Stephen M. Kosslyn
series editor
MIND
The Tempo
in Cons
Benjamin Libet
H A R VA R D U N I V E R S I
Cambridge, Ma
Londo
Copyright © 2004 by the President and Fellows of H
all ri
Printed in the United Sta
Library of Congress Cataloging-in-Pu
Libet, B
Mind time : the temporal factor in consciousness / B
p. cm.—(Perspectives in cognitive
Includes bibliographical referen
ISBN 0-674-0132
1. Consciousness. 2. Time perception. 3. Memory. 4. Cognitive
I. Ti
q
153—dc22
To Ralph Waldo Gerard, Sir John Eccles, and K. Allan C. E
CONTENTS
Foreword
Preface
1. Introduction to the Question

2. The Delay in Our Conscious Sensory Awareness
3. Unconscious and Conscious Mental Functions
4. Intention to Act: Do We Have Free Will?
5. Conscious Mental Field Theory: Explaining
How the Mental Arises from the Physical
6. What Does It All Mean?
Bibliography
Index
FOREWORD
I just typed the word “consciousness” into the sear

in Amazon.com, and it returned 2,670 titles. If I w
weeks, there will probably be more. Does the world r
another book on consciousness? Yes—if we are talking
one you are holding in your hands, right now. This bo
ingly different from most of the others in one key res
cuses on empirical discoveries, not speculation or
Benjamin Libet has an enviable track record of produ
empirical findings about the relationship between neu
and consciousness. And these findings are not simply
they are also surprising. His discoveries were at first
sial, but have withstood the test of time. Surprising fin
a special role in science, given that they (by definition
apple cart of conventional wisdom. His results must n
plained by any theory of consciousness and its neural
nings. This book gathers together Libet’s contributio
place, and puts them in context.
made the decision to act. This finding has major im
for one of the deepest problems in philosophy and
namely the problem of “free will.”
First, a brief overview of the basic discovery: Libet
ple to move their wrist at a time of their choosing. T
pants were asked to look at a moving dot that in
time, and note the precise time when they decided t
wrist. The participants reported having the intention
milliseconds before they actually began to move. Libe
sured the “readiness potential” in the brain, which
by activity recorded from the supplementary mot
the brain (which is involved in controlling movem
readiness potential occurred some 550 milliseconds
action began. The brain events that produced the
thus occurred about 350 milliseconds before the part
aware of having made a decision. Libet shows that th
is not simply due to extra time required to note and
time.
Why is this finding important? Consider two reason
the face of things, the finding suggests that being co
having made a decision might be best thought of as
brain processes that actually do the work, rather th
of the causal chain of events leading up to a decisio
Libet points out that even if a movement were initia
conscious forces, there is nevertheless ample time
act, once one is aware of one’s intentions. Libet b
this observation keeps the door open for traditional
“free will.”
x •
FOREWORD
1. At birth, one’s thoughts, feelings and behavior
mined by genes, prenatal learning, and environmenta
2. Subsequent thoughts, feelings and behavior ar
the foundation present at birth—they are determine
genes, learning history, and present stimuli. All dec
choices are based on reasons, and those reasons are a
sult of one’s accumulation of experience, as modula
netic factors.
3. If one tries to change oneself, both the goals an
of such change are themselves determined by genes
learning, and current environmental stimuli. What on
determined by what one already is.
4. Adding random factors would not confer free
(2002; Stapp, 2001; and others) notes that simply addin
minacy to a system does not make its actions free i
not already free. In fact, adding randomness decrease
rather than increasing it. “Random behavior” is not “f
5. Thus, this argument goes, there’s no free will t
cised during the interval between when one becom
of an impending action and one performs it. Wheth
you will squelch the action is as determined as ar
tors that initiate the action in the first place. Even
time to override one’s unconscious urges, there’s no f
work if one’s conscious decisions are themselves d
(cf. Wegner, 2002). Libet’s “time to veto” no more c
opportunity to exercise of free will than the time bet
ting eggs on the skittle and waiting for them to fry pr
eggs with the opportunity not to cook.
FO
that classical deterministic views are rooted in a w
that is not in fact correct. Many events in the real wo
like pool balls, hitting one another and careening off
the table in predictable ways. We know that many p
tems have chaotic elements: The way they respond t
bation depends on tiny—in principle, never precise
able—differences in their start state. Freeman (2000)
have shown that at least some aspects of brain functi
conceived as such systems. Is it possible that the ver
the brain confers free will? Kane (1996) has suggeste
and I will summarize a version of the type of view h
(although he focuses on process that may occur w
faced with difficult decisions, the basic ideas can b
further).
Let’s consider one possible way in which this fea
brain may keep the door open for Libet’s idea.
1. Libet is right to focus on consciousness when
about free will: In order to employ free will, one mu
information in working memory. Such information i
alternative choices, the rationales for each, and the
consequences of making each choice (although not a
mation must be in working memory at the same tim
ternal force coerces us, or we are operating on “au
lot,” we are not exercising free will.
2. The rationales and anticipated consequences—an
pending on the situation, the alternative courses of
not simply “looked up” in memory, having been sta
like notes in a file after previous encounters. Rathe
xii •
FOREWORD
determined by one’s learning history (even as filtere
genes). By analogy, consider the path of a raindrop
down a pane of glass. It zigs, it zags, tracing a pat
plained with the aid of chaotic principles. The same
striking precisely the same place on that pane on a w
(which would cause the glass to be in a slightly diffe
would take a different path. In chaotic systems, very sm
ences in start state can produce large differences do
The pane of glass is like the state of the brain at any in
pending on what one was just thinking about, the br
different “start state” (i.e., different information is pa
vated, different associations are primed) when one con
tionales and anticipated consequences—which will a
one decides. (Note that this idea does not simply move
lem back a step: What one was just thinking itself w
a result of nondeterministic processes.) Our though
and behavior are not determined; we can have novel
well as “second thoughts.”
3. Given the choices, rationales, and anticipated con
one decides what do on the basis of “what one is”
speaking, to use Strawson’s term, which includes on
edge, goals, values, and beliefs). “What one is” consi
of information in memory, which plays a key role i
cesses that construct the alternatives, rationales, and a
consequences. In addition, “what one is” governs ho
tually makes the decisions. And making that decision a
encing the actual consequences in turn modifies “wh
which then affects both how one constructs alternat
FOR
We are not simply accumulators of environmenta
tered by our genetic make-ups. We bring something
unique to each situation—ourselves. Nietzsche (1886
in Strawson, 1994, p. 15) commented, “The causa sui
self-contradiction that has been conceived so far.” Ma
4. This brings us back to the implications of Libet’
and suggests a way in which we can exercise free
that crucial interval between when we become aware
tion and the action begins: The sum of “what one is
to make a specific decision. Such a decision can oc
sciously, and initiate an action. However, upon realizi
is about to perform a specific act, one can consider its
sequences and the rationales pro and con for perfo
act; this information is constructed on the spot, and
ent during unconscious processing. And, based on “w
one then can decide not to move ahead—or, if the ac
gun, one can decide to squelch it (and thus one is no
the 200 milliseconds Libet has measured). As Libet no
in fact veto an action, and that decision is not a forego
sion. We make decisions for reasons, and those reas
reasons.
Libet has made a fundamental discovery. If the
mental events is as he describes, then we not only
will” in principle—but we also have the opportunity
that free will.
The ideas I’ve briefly sketched are variants of many
Kane, 1996), and address issues that have been discus
times heatedly) for thousands of years. I’ve not men
issue of “ultimate responsibility”—whether one is
xiv •
FOREWORD
leads us to take a step back, and reframe the question:
asking whether one is “ultimately responsible” for ev
of what one is, why not ask whether one is “proxima
sible” for the effects of every aspect of what one is on
does? Can we choose—based on what we’ve chos
come—to override some impulses and express others?
I hope these brief reflections have conveyed two
points. The first is that these are extraordinarily kno
and the question of the role of consciousness in free
likely to be resolved soon. And the second is that we
ing a new era in discussing such questions. No longer
stricted to the arm chair and the silver tongue. We now
jective data. This book makes a crucial contribution in
grist for the mill of anyone interested in consciousnes
responsibility, or the relation of mind and body.
I hope you enjoy reading this book as much as I did
S. M. Kosslyn
References
Freeman, W. J. 2000. Brain dynamics: Brain chaos and intentionali
grative Neuroscience. Bringing Together Biological, Psychological
Models of the Human Brain, ed. E. Gordon. Sydney, Australia
Academic Publishers, pp. 163–171.
Kane, R. 1996. The Significance of Free Will. New York: Oxford Uni
Klein, S. 2002. Libet’s research on the timing of conscious intentio

commentary. Consciousness and Cognition 11:273–279.
FO
Studies 75: 5–24.
Wegner, D. M. 2002. The Illusion of Conscious Will. Cambridge, M

Press.
xvi •
FOREWORD
P R E FA C E
How did I come to write this book?

We had made some surprising discoveries of how t
involved in the production of conscious subjective
and of unconscious mental functions. Where and
scious experience arises, and how that differs from un
mental activities, are questions of profound interest n
me but also to many others. Our discoveries were arr
perimentally. They were based not on speculative theo
on factual findings. This is in contrast to most writing
posals by philosophers and by some neuroscientists,
and others on these subjects.
I thought, therefore, that our discoveries and the ma
tant implications that they generate should be made a
a wide general audience as well as to philosophers,
and clinicians who deal with mental problems. An esp
portant feature of this presentation is the demonstr
experience can be studied only in awake human su
can give you a report of their experience. Non-hum
may very well have conscious experiences, but t
good way of studying those experiences validly. I
the unique opportunity to study human subjects
undergoing neurosurgical therapy from Dr. Bertram
Bert and I were formerly colleagues in the Biomecha
UCSF, where he was a neurologist. After three years
in neurosurgery in Sweden, Bert started a practice
Zion Hospital in San Francisco. He also wanted to
portunities of access to the human brain to conduct
risk-free investigations, and he offered me the opp
conduct such studies. The therapy required the pla
electrodes at specified structures within the brain. I
the chance of studying the electrical activity of cer
cells and electrical stimulation of appropriate nerve c
tion to reports of conscious experiences by patient
emphasize that our experimental procedures added
the patients. They were done with the informed con
patients and not one of our activities produced any di
damage. The patients were in fact remarkably coope
our studies.
Dr. Feinstein was himself easygoing and coopera
operating room. He let me design the experiments,
not display an autocratic prima donna attitude duri
After almost twenty years of this collaboration, Fei
prematurely in 1978. Following his untimely deat
turned to the study of voluntary action, in which w
xviii •
P R E FA C E
nal (Libet et al., 1991). For this latter study, we had av
tients with permanently implanted stimulating electr
sensory pathway in the brain, for the relief of intrac
These patients were made available to us by the coop
Dr. Y. Hosobuchi and Dr. N. M. Barbaro, neurosu
UCSF. That work was also made possible by Michael M
Professor of Physiology, who generously provided
space and computer equipment for our use at UCSF.
All these studies began in 1959, with the added col
of W. Watson Alberts as a biophysicist, and Elwood (
Wright, a biomedical engineer. Watson left the gro
to become a successful administrator at the Nationa
of Nervous and Mental Diseases. He was replaced
Gleason, a bioelectric engineer. I owe much of our ef
to the contributions of this team of collaborators. I
express our appreciation to the many patients who c
in the studies. In addition, a group of ten graduate s
psychology were enthusiastic subjects in our experim
ies of voluntary action and conscious intention to act.
The three neuroscientists to whom I am dedicating
were the chief mentors in my scientific career. Ra
Gerard, starting with my graduate work at the Un
Chicago, introduced me to imaginative research in ne
and maintained his faith in my abilities even during m
riod. Sir John Eccles brought me into modern exp
neuroscience (during a year-long period of research
tion at the Australian National University) and supp
working on mind-brain relationships, even when that
P
the brain at the Institute of the Pennsylvania Hospit
delphia.
I am grateful to my grandson Victor Libet and my
Gayla and Moreen Libet, for their helpful comm
readers of early versions of the manuscript. I also
friends Robert Doty and Anders Lundberg for their v
vice and continuous encouragement and support. Co
Michael Fisher, science editor for the Harvard Unive
led to a major reorganization of the coverage in the b
beth Collins provided skilled editing. I am grateful
Kosslyn for his excellent and meaningful Foreword.
Finally, I benefited from my wife Fay, my childr
Moreen, Ralph, and Gayla), and my grandchildr
Anna, Leah, Lev, and Stavit).
xx •
P R E FA C E
M
1
INTRODUCTION TO THE QUESTION
Something happens when to a certain brain sta

certain “consciousness” corresponds. A genuine glim
into what it is would be the scientific achievement be
which all past achievements would pale.
—William James (1899)
You stop to admire the intense blue of a flower; you

with the antics of a young child; you feel pain in a
shoulder; you listen to the majestic music of Hande
and feel moved by its majesty; you feel sad about a
ness; you feel you can make a free volitional choic
to do about a job and when to do it; you are awar
thoughts, beliefs, and inspirations; you are aware of
self as a real and reactive being.
All of these feelings and awarenesses are part of yo
tive inner life. They are subjective in the sense that th
cessible only to the individual subject who is experien
They are not evident in and cannot be described by ob
of the physical brain.
the physical brain is essential to and intimately invo
manifestations of our conscious, subjective experienc
That fact gives rise to some fundamentally impo
tions.
The Problem: Relating Brain

Activities to Mental Functions
Appropriate nerve cell activities can certainly influen
tent, or even the existence, of subjective experience
verse true? That is, can our conscious intentions reall
or direct the nerve cell activities in the performance
voluntary act?
Our subjective experiences are based on widesprea
of thousands of nerve cells, located in separate pl
brain. How is it possible for our experience, like a vi
to appear subjectively in a unified form?
There is a further important issue when consid
scious experience. Many of our mental functions are
unconsciously, without conscious awareness. The cons
perimental and clinical evidence for that assertion is
later chapters. The role of unconscious mental proce
emotional existence was, of course, prominently de
Sigmund Freud and others. The question in the con
present interest becomes, How does the brain dist
tween conscious and unconscious mental events?
Finally, there is the most mysterious of these
How can the physical activities of nerve cells in the
2 •
ness, and so on? How can the gap between the “phy
brain) and the “mental” (our conscious, subjective ex
be bridged?
There have been many proposed answers to these
questions (for example, see Hook, 1960). These h
mainly from philosophical and religious sources, alth
tributions from neuroscientists have begun to appea
years. Religious proposals are clearly metaphysical b
scientifically testable. Those from philosophers h
largely theoretical speculative models that are mostly u
As the philosopher of science Karl Popper (1992) po
if a proposal or hypothesis cannot be tested in a way
potentially falsify the proposal, then the proposer can
view without the possibility of its being contradicte
case, a proposal can offer any view without being
Proposals that are untestable in that sense have been
only by philosophers and theologians but even by so
tists. Many scientists like to think their own exp
research—for example, in immunology or motor cont
oretical physics and cosmology—provides a basis for
speculations on the nature of conscious experienc
brain. Although often interesting, these speculations
untestable. However, some of these proposals provi
tive scientific approaches to the problems, and som
philosophical analyses help to define the nature of the
and some limitations on the kinds of answers one ca
achieve.
It is not the intention of this book to present a full
INTRODUCTION TO THE Q
produced direct discoveries with fundamental implic
these form the major coverage in the book. Our
physiological observations were directly related to
conscious experiences by awake human subjects. Tha
has been relatively unique in this field of interest. Rela
mental studies and philosophical views are discussed,
vant and desirable for facilitating the reader’s under
our studies. (For a general history of discoveries in
brain, see Marshall and Magoun, 1998.)
General Views on Mind and Matter

At one pole is the determinist materialist position.
losophy, observable matter is the only reality and
including thought, will, and feeling, can be explain
terms of matter and the natural laws that govern m
eminent scientist Francis Crick (codiscoverer of the g
lecular code) states this view elegantly (Crick and K
“You, your joys and your sorrows, your memories an
bitions, your sense of personal identity and free will
no more than the behavior of a vast assembly of ner
their associated molecules. As Lewis Carroll’s Alice
phrased it: ‘You’re nothing but a pack of neurons (ne
According to this determinist view, your awareness
and the world around you is simply the by-product
nomenon of neuronal activities, with no independen
affect or control neuronal activities.
4 •
tests. It is true that scientific discoveries have increas
duced powerful evidence for the ways in which ment
and even the nature of one’s personality, are depende
can be controlled by, specific structures and functio
brain. However, the nonphysical nature of subjective
including the feelings of spirituality, creativity, cons
and imagination, is not describable or explainable dire
physical evidence alone.
As a neuroscientist investigating these issues for m
thirty years, I can say that these subjective phenome
predictable by knowledge of neuronal function. This
trast to my earlier views as a young scientist, when I
the validity of determinist materialism. That was befo
my research on brain processes in conscious experien
40. There is no guarantee that the phenomenon of
and its concomitants will be explainable in terms of
known physics.
In fact, conscious mental phenomena are not redu
explicable by knowledge of nerve cell activities. You
into the brain and see nerve cell interconnections a
messages popping about in immense profusion. But y
not observe any conscious mental subjective phenom
a report by the individual who is experiencing such p
could tell you about them.
Francis Crick demonstrated his scientific credential
ing his physicalist-determinist view an “astonishing h
awaiting future developments that might produce
Actually, even the nonmental physical world exhi
tainties (quantum theory) as well as chaotic behaviors
a deterministic predictability of events impossible.
conference on these issues, the eminent theoretical p
gene Wigner was asked whether physics could ever e
sciousness. Wigner replied, “Physics can’t even expla
let alone consciousness! The more meaningful ques
fore, would be: Does the phenomenon of conscious
and its relation to the physical brain, fully obey the k
and laws of the physical world? (More on this later.)
At the opposite pole from determinist materialism
that the mind is separable from the brain (dualism).
version of dualism may maintain a belief in the exi
soul that is somehow part of the body during life, bu
rate and take off to variously defined destinations of
ity after death.
I shall state, at once, that the latter is absolutely t
belief. The same is true for most other philosophic
gious proposals. There is nothing in all of scientifi
that directly contradicts such beliefs. Indeed, they
within the purview of scientific knowledge (see Ka
position, described earlier).
A beautiful example of the scientific process was gi
stein’s proposal that light is subject to the same grav
fluences as matter. However, to demonstrate the g
effect on light requires that the light pass near an ob
mense mass, one far greater than that available on
difficulty in providing a proper test prevented full acc
6 •
ble during the eclipse. Indeed, the star’s apparent po
altered, as the light was bent from its path by the “p
sun. Had the light not been bent, Einstein’s proposal w
been falsified (contradicted).
Is There Any Scientific

Approach to the Mind-Brain Problem?
Is there some way to arrive at convincing knowledg
conscious subjective experience arises? Is there a way
that is based on observable evidence?
We must first recognize that the brain is the physic
for conscious and unconscious mental functions. For
know it, the necessity of the appropriate function and
of the brain is incontrovertible. There is no objectiv
for the existence of conscious phenomena apart from
(A belief in a separable conscious soul is not excluded
previously.) Perhaps the most convincing piece of evi
it is the brain and not any other bodily structure tha
lies in the effects of a complete severing (transaction)
nal cord at its junction with the brain. This unfortu
occurs not infrequently in accidents in which the ne
ken,” as in the recently publicized case of the actor C
Reeves. The patient remains the same conscious person
fore the accident. However, he loses all control of b
ments from the neck down, including of breathing m
as well as all sensations that are carried by spinal ner
body. Interruption of the nerve pathways that connec
to the head. And, if the brain is functional, the per
awareness of his thoughts, feelings, and self.
On the other hand, damage to the brain itself can r
loss of various conscious functions, or even a perman
consciousness, depending on the sites of the damag
loss of brain function that truly defines the end of co
man life, that is, death. This is so even when the rest o
including the spinal cord, skeletal muscles, and the he
functioning. Indeed, under this condition of brain
other organs or tissues may be taken for transplantati
people.
In earlier times, the heart was often regarded as
consciousness and of emotional feelings (see Aristot
placing one heart with another (even one that is a
device) does not alter an individual’s emotional make
rience.
So, what sorts of factual answers to the questions
scious experience could we hope to pursue succe
what answers have we now achieved? One important
how brain activities are related to conscious and u
mental function—is, in principle, amenable to desc
experimental investigation. But to do that, we nee
conscious subjective experience, and do so in a way t
ational—that is, practical for study.
We start with the stubborn fact that a conscious su
perience is directly accessible only to the individual w
experience. Consequently, the only valid evidence fo
nal observer must come from an introspective repor
perience by the subject.
8 •
Scientists, like philosophers, have speculated about
brain and mind are connected. But, until recently,
including neuroscientists, have attempted direct exp
studies of how cerebral nerve cell activities are invol
production or appearance of conscious, subjective ex
Why? Apart from the technical difficulties for such ex
on human subjects, a philosophical impediment ha
major role.
Studies that require data from introspective reports
tive experiences have tended to be taboo in the acad
munity. That negative attitude was influenced in lar
the dominance, during the first seventy-five years of t
eth century, of behaviorism in psychology and of logi
ism in philosophy. These viewpoints hold that only d
servable events are admissible as scientific data. Int
reports are only indirectly related to the actual subject
ences; that is, they are reports of something not direc
able by the investigator and are untrustworthy ob
However, unless scientists can find a way to obtain v
spective reports, they can never study the profoundly
question of how our conscious mind is related to our
late great physicist Richard Feynman stated, “I’m just
find out more about the world . . . Whatever way it c
it’s nature, and she is going to come out the way she
fore, when we go to investigate it we shouldn’t pre-de
it is we’re going to find.”
We must, of course, admit that an introspective re
not provide absolute evidence about the experienc
is one’s own experience—as noted by René Descar
Berkeley, and others. Yet, in our ordinary social inter
commonly accept introspective reports of experienc
individuals as meaningful reflections of their expe
though we may try to evaluate the validity of these r
To be sure, the conversion and transmission of an
into a report may involve some distortion. However
ble to limit the kinds of experiences being studied to
ones that do not have emotional content. These expe
even be tested for reliability. In our own investigatio
very simple sensory experiences that had no emotio
that might lead to distortion. Furthermore, we could
liability of the reports, by changing the sensory inp
under the investigator’s control and comparing the d
ports elicited in this way. It should have been clear
that a way to study subjective experiences scientific
achieved.
I should add that an introspective report need not b
a verbal, oral statement. A nonverbal report, like
appropriate key to indicate whether a sensation had
jectively felt, can be quite acceptable, providing the
derstands that this indicator in fact refers to a subjec
spective experience.
I may add here that when I was an undergraduate
that verbal expressions are not completely adequate
tions of reality. They are only approximations, lim
meanings attributable to the words. I decided, there
to think about reality in a nonverbal way—that is, to t
the real situation in a fully integrated and intuitive
10 •
became a major factor in shifting scientific opinion o
fulness of introspective reports. Cognitive scientists
deal with questions about what people knew and felt
that was related to reality. To do so, they had to have
them about their subjective experiences. I should note
are still traditional behaviorists among psychologists,
large group of philosophers adhere to a movement
behaviorism called functionalism.
Starting in the late 1950s, I did not wait for cognitive
support my use of introspective reports in our stu
proached this issue as a physiologist, with no stake in
ism or functionalism. My attitude was, from the start
scious experience could be studied and treated like
observable function of the brain. As an experimental
was, and is, my firm conviction that a person’s repor
scious experience should be regarded as primary evid
evidence should not be altered or distorted so as to b
conform to a preconceived view or theory about the
consciousness. Unless they can be convincingly affect
tradicted by other evidence, properly obtained intros
ports of conscious experience should be looked on
kinds of objective evidence.
I was, in fact surprised when I found that a contro
of opinion among behavioral scientists did not agre
view. Indeed, a visiting group of such individuals, repr
study section of the National Institutes of Health, told
not studying a suitable topic. They denied my applic
grant.
Interestingly, I found no such rejection among th
INTRODUCTION TO THE QU
berg, Robert Doty, and Howard Shevrin. These rese
garded our work as praiseworthy and pioneering—
also expressed during a major symposium entitled
Conscious Experience” in 1964. Sponsored by the Pon
demy of Sciences and chaired by Sir John Eccles, t
sium was held in the fifteenth-century house of Piu
the Vatican grounds. Pope Paul took us seriously enou
a formal audience with us. The twenty-five or so m
the symposium were seated on one side of a great
roughly equal number of Cardinals faced us on the o
their red robes. When the Pope came down to gr
Catholic scientists knelt and kissed his ring, and the
shook his hand. I still have the thick red leather nam
gold lettering from that meeting. Since then, I have
ticipant and speaker in a number of additional inter
posia on consciousness. There was, in fact, another
Vatican in 1988, again organized by Sir John Eccles.
Besides neurophysiologists, leading philosophers s
late Sir Karl Popper, Thomas Nagel, and the late Step
also agree with my views concerning how to study
subjective experience. Stephen Pepper was Professor
ophy at the University of California–Berkeley. Pep
strong advocate of so-called identity theory, which ho
externally observable physical quality of the brain an
quality of subjective experience are simply different
nological aspects of a single “substrate.” Neverthel
listened carefully to my discussion of my team’s view
ings; he even concluded that our evidence for a retroa
12 •
Awareness
The use of introspective reports in our studies led me
nize the nature of their significance more clearly. I re
the essential feature of introspective reports of conscio
ences is awareness, or being aware of something. W
aware of encompasses a great variety of experientia
including awareness of the external world and of ou
bodily world (via sensory inputs), of our feelings (ang
pression), of our thoughts and imaginations, and of o
Many, if not most, philosophers have spoken o
kinds and levels of conscious experience. Commonly,
ness is regarded as a special case and one that may be
human beings and possibly chimpanzees. We canno
that the experiential contents even of similar events ar
in other people. For example, what I see as yellow
be identical to what you see as yellow, even though
learned to give that kind of experience the same nam
be much more confident that the feature of awarene
the other person, is fundamentally identical to our o
ness, even if the contents of that yellow experience m
identical.
I suggest, therefore, that there is no need to inven
kinds or categories of consciousness or of conscious e
to deal with all the kinds of experiences. The comm
in all cases is awareness. The differences lie in the dif
tents of awareness. As I will argue from the experim
dence, awareness per se is a unique phenomenon, an
been called qualia by philosophers. Such experience
phenomena not explainable by the physical nature of
that produce them or by the corresponding neural ac
so they pose difficulties for materialist theories abou
experience. But I see no reason for setting up these
problem that is fundamentally different from other aw
Awarenesses of all kinds are equally unexplainable by
theories.
We should distinguish between “conscious experi
the state of simply being awake and responsive—in o
being in a “conscious” state. To be in a conscious
course, a prerequisite for the appearance of consciou
experiences, except in the case of dreams. In dreamin
conscious experiences during the sleep state. Howeve
of being awake and conscious, and the phase of sl
which dreams appear, both require a diffuse activatio
rebral cortex by structures in the brain stem and in th
(the structure at the base of the forebrain, below t
hemisphere). This function in the brain is a nece
ground condition for the production of conscious ex
How Can We Study the Relation between the

Brain and Conscious Subjective Experience?
My attitude has always been the same: never mind t
tive untested theories. Rather, focus on finding ou
brain actually deals with or brings about the appea
conscious experience. I suppose this attitude stem
background in experimental neurophysiology. Our go
14 •
man subjects—with whom direct studies of brain fun
of course, severely limited. Much can be done with
a behavioral level for work on memory and learning, r
tion of visual processes (spatial, colors), and so on.
such functions can be performed without conscious
as they can be even in human subjects. As Mari
Dawkins put it, One should take care not to “blur the
distinction between being clever and being conscious
we should be on guard against “the impression that a
to do to probe animal (or human) minds is to show th
pable of various complex intellectual tasks and we w
bly have shown that they are conscious” (in other wor
tively aware).
Only recently has an experimental design been d
Cowey and Stoerig, 1995) that can generate some
that a monkey is employing conscious awareness in or
form a subtle task. This design involved monkeys with
the primary visual cortex. The same lesions in human
a loss of conscious vision, or blindness. Visual stimuli
sented and the monkeys’ ability to detect these st
tested. When a monkey had to give a forced choice
(yes or no), stimuli presented in the defective “blind”
were detected 100 percent of the time. Human patie
similar defect in the visual cortex can point correctly t
though they claim they cannot see it (a phenomen
“blindsight”). But when a monkey was allowed to resp
it classified such stimuli in the defective visual field a
in other words, nothing there. In this condition, the m
scious vision and unconscious detection.
For our experimental approach to the question
study the relationship between the brain and consci
tive experience, I set out two epistemological princ
believe must be followed: the introspective report as
tional criterion and no a priori rules for mind-brain re
1. Introspective report as the operational criterion. I h

discussed the case for introspective reports. Here is an
corollary of this principle: Any behavioral evidence
not require a convincing introspective report cannot b
to be an indicator of conscious subjective experien
so regardless of the purposeful nature of the action
complexity of cognitive and abstract problem-solving
both can and often do proceed unconsciously withou
by the subject. One must even be careful to distingui
the ability to detect a signal and the awareness of the
Behavioral actions are the observable muscle actio
tonomic changes (in heart rate, blood pressure, sweat
on). Purely behavioral actions that are not reportin
spective experience cannot provide valid evidence of
subjective experience. In reporting an introspective
the subject is responding to a question about her priv
ence and we are confident that she understands the
behavioral act made without this condition may, in f
formed unconsciously.
2. No a priori rules for mind-brain relationship. A coro

tion is, Can one describe what a person is feeling or t
16 •
brain and observe the multifarious activities of nerve
various structures, you would see nothing that looks l
tal or conscious phenomenon. This point was alread
the great seventeenth-century philosopher and math
Leibniz, among others.
In contrast, another great mathematician, Laplac
enamored of the new mechanistic models in Newton
Laplace proposed that if he could know all the pos
states of energy or motion of all the molecules in the
he could predict all future events. He argued that k
of all such molecular features in the brain would e
to specify and predict what was going on mentally.
proposition is not testable in practice. Not only ca
hope to specify the data for the astronomically large n
molecules in the brain, but modern physics tells us
impossible, in principle, to measure the position an
simultaneously for any particle. Second, even if we co
Laplace’s condition, we would see only molecular
tions, not any mental phenomena. It is curious that a
group of philosophers, the functionalists, still hold a
istic, Laplacean-like view.
The general principle to be followed, in contrast to
ism, is that externally observable “physical” events and
observable “mental” events are phenomenologically
dent categories. The two are certainly interrelated, bu
tionship between them can be discovered only by sim
observations of the two separate phenomena. The re
cannot be predicted a priori. Neither phenomenon is red
describable by the other.
Instead, he reports feeling something in a part of hi
the hand, even though nothing actually occurred in
An external observer would have no way of describin
jective experience without asking the subject for an in
report about it.
This principle leads to a flat rejection of the reduct
popular with many scientists and philosophers (fo
Churchland, 1981; Dennett and Kinsbourne, 1992). A
this view, knowledge of the neuronal structures an
(or their molecular underpinnings) is sufficient for d
explaining consciousness and mental activities. But w
seen how that reductionist view would not work.
Where in the Brain Are the

Processes Related to Conscious Experience?
The eminent neurosurgeon Wilder Penfield, with his
(especially Herbert Jasper), made thousands of obse
patients’ introspective reports in response to local elec
uli to the cerebral cortex. (The cortex was exposed
during therapeutic procedures to identify foci of ep
zures. Patients were awake, with local anesthetics ap
scalp.) Other neurosurgeons also carried out such m
responses. Reports of sensations were obtained by
the primary sensory areas of the cerebral cortex, whe
tosensory (bodily sensations), visual, or auditory. Cle
cal stimulation of the primary sensory cortex would
place to study the requirements for producing a co
portable event.
18 •
sponses (direct cortical responses, or DCRs) are recor
all stimulus sites. Presumably, the production of a
conscious response in silent areas may require more co
tivations than are possible at the primary sensory ar
cortex. Or, silent areas may not mediate conscious fun
In any case, it is worth emphasizing, from this and
dence, that considerable amounts of neuronal activity
without eliciting any conscious experience.
Penfield and Jasper were impressed by the observ
large destructive lesions in the cerebral hemispheres
sult in loss of consciousness, while small lesions in the
systems, located in the brain stem or in the intralam
of the thalamus, did produce a loss of consciousnes
Penfield (1958) therefore proposed that the “seat” of
ness is located in these medially situated subcortical
which he called the centrencephalic system. Similar
eminent neurosurgeon, Joseph Bogen (1995), recently
that the conscious function resides in the intralamin
of the thalamus, a component of the centrencephalic
The logical difficulty with Penfield’s and Bogen’s v
it does not distinguish between necessary and suffici
tions. That is, even if a structure is necessary to the
function, that does not, in itself, make that structure a
condition for producing conscious experience. There
many other functions that are necessary for the possib
tion of conscious experience. For example, if the h
beating, a person loses all conscious functions within
onds. But the heart is not where conscious functions r
trary to the notions of many earlier peoples; one can
the view that specific kinds of neuronal activities in t
hemispheres underlie production of conscious events
A variety of studies have given us important i
about where in the brain there are nerve cell activ
ated with conscious or behavioral events. These fa
groups: neuropsychological studies and techniques f
ing changes in brain nerve cell activities.
1. Neuropsychological studies have examined chang
tal functions produced by destructive lesions in speci
calities. The “granddaddy” in this field is perhaps t
Phineas Gage, who sustained accidental damage to a
tion of both cerebral hemispheres. As he was layi
tracks, a metal rod was accidentally propelled into o
Gage’s head, in front of the temporal lobe; the rod
through to the other side of his head. Gage surviv
personality changed dramatically. Previously a stab
and sociable person, he became uninhibited (swearing
easily aroused emotionally), unreliable in his work, an
in foresight and planning. His case highlights the imp
the frontal lobes of the brain in functions of self-co
ning, and so on. Much more has since been discov
functions of the frontal lobes.
More recently, neuropsychologists have been deve
resentations of subtle differences in mental funct
which we had no previous inklings. For example, cert
lesions produced by local damage from blood clots
(in other words, small strokes) can produce the inab
tect consonants in spoken words although the vow
20 •
tensity of nerve cell activities in the brain. These tech
based on the premise that an increase in local neural
accompanied by an increase in the energy metaboli
nerve cells. That increase in metabolism could result
local consumption of oxygen and the release of ce
products of metabolism into the local spaces around t
cells. Most notably, oxidation of glucose produces ca
ide (CO2). CO2 is known to produce a dilation of the
rioles, thereby increasing the circulating blood in that
The first successful technique for measuring chang
or regional cerebral blood flow (rCBF) was devised by
ish clinical neurophysiologist David Ingvar and his
(see Lassen and Ingvar, 1961; Ingvar, 1979, 1999). In pri
technique involves measuring and mapping local cha
dioactivity, after an injection of a relatively safe dose
active compound into the cerebral blood supply. A larg
of scintillation counters are arranged on the subject’s
of these counts the radioactive emissions at its site an
cords the degree of radioactivity of the injected comp
locality. An increase in the radioactivity at certain
times after the injection indicates an increase in the
blood carrying it into that region.
Ingvar and his colleagues studied changes in rCBF
with sensory inputs and motor activities but also wit
processes. They found, for example, that if a subj
imagined moving her fingers, without actually moving
creases in rCBF could be detected in some of the s
that “lit up” when subjects moved their fingers volun
activations of sensory or motor areas.
One could argue that such results are evidence of
of a conscious mental process to influence neuro
in the brain. One could also study pathological co
look for local or general abnormalities in cerebral
whether in the resting state or in response to relevan
tasks. For example, certain deficiencies in blood cir
the brain have been observed in people with early A
disease, schizophrenia, and other health issues.
Louis Sokoloff and his team (1977) pioneered imp
surement of local metabolic changes in the intact b
work led directly to the invention of more powerful
detect changes in metabolism. As with the Ingvar
no surgical penetrations into the brain were require
methods could be used in human subjects. The tw
widely in use at present improved both the spatial
and the speed of measurement.
The first of these methods is positron-emission-to
(PET scans). This method involves injecting weakly
substances that emit positrons, instead of electromag
tions. The positrons are detected by a large number o
vices arranged on the scalp.
The second method uses magnetic resonance ima
to show quantitative changes in a variety of atoms (s
gen and carbon) associated with the neural functions,
localized fashion.
All of these investigations—studies in neuro
rCBFs, PET scans, functional MRIs—give us inform
about where in the brain the nerve cell activities may
22 •
the timing in the relation between changes in nerve
ties and a mental function (such as the relation betwee
ing of a change in brain activity relative to conscious
of a given event). Indeed, it is even possible that the a
ing increased activity are not the sites of primary imp
the initiation or organization of the functions being t
primary sites could be smaller and show much weak
in the measured images.
Even when one of these methods becomes capab
fast resolution of changes in time, as the functional
become, estimation of the timing of the neural chan
ited by the physiological process being measured.
method (like the PET scan) measures a change in
culation of blood or in chemical constituents prod
metabolic change in the nerve cells. Such metabol
mostly follow the functionally relevant changes in the
(whether these are synaptic responses or alterations i
nerve impulses). The important relevant changes in
the nerve cells can occur in milliseconds; but the me
ergy changes, initiated by these neural activities, ma
onds to produce the changes that are measurable by t
niques. Thus, it is not possible to answer questions suc
conscious intention precede or follow the cerebral init
voluntary act?
Electrophysiology
The recording of electrical events that are an intrins
nent of the relevant neuronal activities does allow us
nerve impulses (by their action potentials), and by the
nonconducting synaptic potentials. Synaptic potentia
duced when fibers coming from other nerve cells de
coming message to their terminals, where they ma
with specialized areas of the surface membrane o
nerve cell. The specialized junction between the inco
and the site where it terminates on another cell is c
apse(from the Greek for clasping of two elements). In
apses the incoming terminal can release a special c
neurotransmitter. The area of cell membrane on th
side of the synapse contains receptors specialized to
the neurotransmitter.
The postsynaptic response usually results in a loc
change, making the external side of the receiving me
ther more negative (with excitatory effects), or mo
(with inhibitory effects). In either case, a difference i
potential (voltage) is thus created between the local p
portion of cell membrane and the adjacent membran
ilarly affected) of that same cell. That produces a field
current around the cell. The voltage changes in that e
can best be detected by an electrode placed in the ex
dium close to the cell. However, smaller voltages o
can be recorded at greater distances, using suitable am
Thus, very small voltages, in the microvolt range,
tected even on the scalp. These are the basis of the
cephalogram (EEG; also called the “brain waves”). T
electrical rhythms were first reported in humans by H
in 1929. The EEG is now widely used in studies of
24 •
accompany epileptic activity are used to diagnose and
site of epileptic foci.
Magnetoencephalograms (MEGs) have also come i
cently. These are recordings of the small magnetic fie
ated by electrical currents. MEGs, recorded with detec
scalp, are claimed to give better indicators of the o
neural fields than do EEGs. However, small electro
duced intracranially to make direct contact with the c
face or inserted into deeper subcortical structures
electrical changes that are more localized and more m
than any scalp recording, whether EEG or MEG.
It is possible, and often desirable, to perform som
surgical procedures without general anesthesia in an
tient. To accomplish this, a local anesthetic is injecte
scalp and tissues covering the cranial bone. That pr
sufficient to block any pain, as no pain is generated b
hole in the cranial bone or by making contacts with
sues. There are no special nerve endings that respond
in the brain, such as those that respond to injury else
that lead to feeling pain when these messages rea
places in the brain. Pain is very important for inform
tissue injury, so that we can try to get away from the
injury. Presumably, there is no adaptive value for such
system in the brain itself. The brain is encased in a
bony cranium. Any object that could produce injury to
would first produce pain when it penetrates the scalp a
of the bone, as well as the membranes (meninges) tha
brain. However, an injurious tumor growing within
of the first to show that electrical stimulation of the a
sensory cortex could evoke reports of tingling sensat
subject. (This was before the days of electrical record
When the stimulating electrode was placed on the
gyrus (the rounded surface lying behind the central
fissure of the cerebral cortex), a tingling sensation
some part of the body. It was not felt at the stimula
the brain. Stimulation of the precentral gyrus, just
the fissure, produced local movements of various p
body. These areas then comprised the primary som
area and primary motor area of the cerebral cortex.
Somewhat later, the German neurosurgeon Otfr
(referenced in Penfield, 1958) and the American-Cana
surgeon Wilder Penfield greatly expanded the kno
tainable this way (see Penfield and Boldrey, 1937).
vassed virtually the whole cortical surface, in variou
They, and others since, found that stimulation of mo
the cortex produced no reportable sensations, mov
feelings; these areas were called “silent.” The “excit
that did produce responses were limited to the so-call
sensory areas for bodily and somatic sensations, the
(in a defined area at the occipital, rear pole of the c
the auditory cortex (on the upper forward area of th
lobe). Penfield also observed psychic reports of hal
memories, and so on when he stimulated some ar
temporal lobe. (Indeed the temporal lobe with its
structures—hippocampus and amygdala—is now reg
important mediator of memory formation and ce
tional feelings, particularly of fear and aggression.)
26 •
the subject? I proposed that the crude excitation of loc
of nerve fibers would be unlikely to lead to the kin
nized activities that may be necessary to activate a sub
perience. Indeed we are lucky that stimulation of the
sensory and motor areas can elicit subjective response
only guess that happens because the nerve fibers excite
localities have sufficient access to nerve cells directly
these subjective responses.
On the other hand, electrical stimulation of the sile
animals (cats, monkeys) has been shown (by Robert
others; see Doty, 1969), to participate as part of a co
reflex (CR). In an ordinary CR, an effective unconditi
ulus (US) produces a natural response that requires no
For example, a mild shock to a paw (the US) causes
to withdraw its paw. If an unrelated conditional stimu
applied in less than 1 sec before the US, the anima
withdraw the paw when that CS (say an auditory tone
alone. An electrical stimulus to silent cortex can act li
conventional CS, equivalent to sounding a tone. That
mal can learn to withdraw its paw when that cortica
was given alone. This and other evidence indicates
cific activation of neurons in almost any part of the ce
tex can be detected by a subject in a functionally effe
ner. Such detection of electrically activated neural re
the silent cerebral cortex is presumably made uncon
one extrapolates from the absence of any conscious e
with similar stimulations in human subjects.
Whether a stimulus to silent cortex in a human s
also be detected unconsciously is an interesting
the sensory pathway, even when inadequate to produ
scious experience, can nevertheless be usefully detec
human subject (Libet et al., 1991; see Chap. 4). The im
ference is, then, that neural activities inadequate to p
subjective experience or awareness can nevertheless h
diate functions without awareness. Indeed, much o
activities are of that nature.
Our Experimental Entry

My opportunity to get into such studies came fro
league and friend Dr. Bertram Feinstein. Bert was an
tal neurologist with the Biomechanics Lab at UCSF. I
ated with him there in work on muscle functions
locomotion. Bert converted to neurosurgery by spen
years of study in the early 1950s, with the great neuro
Sweden, Lars Leksell.
He then introduced stereotaxic neurosurgery to
cisco, actually to the western United States (see Fein
1960). In stereotaxic neurosurgery, a therapeutic e
probe is introduced into the brain, so as to reach a
deeper structure without cutting the brain open to g
frame with coordinates in three dimensions is fixed
tient’s skull. The coordinates for reading the target i
are mapped in advance. At that time, the method w
used to inactivate certain deep structures, by a heatin
relieve tremor in Parkinsonian patients.
The Leksell-type frame permitted Feinstein to r
28 •
to back. As a result, he could adopt any angle of entry
given target. He could thus choose a track in which th
shaft might go through other structures of research i
the way to the therapeutic target.
Bert Feinstein was unusual among neurosurgeons in
to use such opportunities to study questions of funda
terest, providing such studies could be done with ess
added risks to the patient (and, of course, with info
sent of the patient and the approval of the hospital c
overseeing human experiments.)
When Feinstein offered me the opportunity to dev
while basic studies for which intracranial access to th
awake subjects was required, I immediately decided
try to find out how activities in the brain are related
duce a conscious experience. This question was one
not be pursued in nonhuman animals, because anim
give introspective reports of subjective experience.
The pursuit of how brain activities relate to or pro
scious experience had been a long-range goal of mine
cinated by the question of how our conscious subject
ence could arise in the brain. I did my graduate schoo
on the electrophysiological activities of the isolated
with the eminent neuroscientist Ralph Gerard, my p
the University of Chicago. Gerard asked me, at one po
my first year with him, to make a list of my views o
spontaneous electrical brain waves were doing for the
item I put down was that these waves might be a neu
sion for consciousness of the frog! Gerard had a b
To facilitate the studies, for both clinical and basic
tal purposes, Feinstein had a new operating room con
the Mt. Zion Hospital in San Francisco. This room
cally shielded and contained conduits for electrical re
nerve cell activities in the brain and for delivering elec
uli. The conduits went to an adjacent control room f
trical equipment and operators of that equipment.
Our studies during the initial several years, beginn
were made during neurosurgical procedures in the
room (see Libet et al., 1964). The patients were awake
a local anesthetic applied to the scalp and to the perio
covering the bone of the skull. Each patient had, of c
viously given an informed consent for the essentia
experimental procedure, which included a provision
tient to terminate the study at any time. Patients wer
remarkably cooperative and consistent in their respo
ever, we were limited to about thirty minutes of stud
erating room. It was, therefore, essential to have the
oughly organized and planned for an efficient and
session. We needed a period of relaxed rest afterwa
off from the concentrated discipline of the procedure
The study sessions became more relaxed and fru
Feinstein altered the therapeutic procedure in the 196
ferred leaving the inserts in the brain for some days
to allow the therapeutic lesions to be made in stage
patient in a more normal, ambulatory state. That
change permitted us to study the patients more ful
more leisurely pace outside the operating room. Late
30 •
tients at length, even during return visits to Dr. Feinst
Feinstein died prematurely in 1978. I lost a dear frie
world lost a pioneer in experimental neurosurgery.
also changed the direction of my research. I turned to
how conscious will is related to brain functions. T
could be carried out with normal subjects. Recording
on the scalp were sufficient for detecting the electric
accompanying a voluntary act, and that was what I n
the experiment I had designed. (See Chap. 4 for a fu
tion.)
Of course, even with the cooperation of a neurosu
Feinstein and of suitable subjects, the number of ava
jects for complete studies was severely limited. But th
even single-case studies can be argued, as discussed b
Marshall (1989) in his review of the book by Tim Sha
Neuropsychology to Mental Structure, 1989). Claude Be
great physiologist of the late 1800s) argued that the us
averages in medicine and physiology “leads necessarily
And Bernard was quoted, in support, on a study i
physiologist “took urine from a railroad station where
all nations passed, and who believed he could thus
analysis of average European urine. . . . If we have le
thing it is that only the fine detail of a patient’s perform
suffices for model building, and that, at this level, th
important individual variation is paramount.”
The rest of this book is oriented around the u
perimental developments and discoveries we were f
able to make on the great fundamental issue: How
coveries.
I hope that you, the reader, will see how we design
perimentally tested hypotheses that were generated
new findings, and that you will be able to share in th
scientific inquiry and in the excitements and thrills o
coveries. In contrast to most other books about con
you are about to be exposed to direct experimental evid
testable theories on this issue, rather than to speculative
nontested constructions.
32 •
2
THE DELAY IN OUR CONSCIOUS SENSORY AWAR
If you tap your finger on a table, you experience the e

curring in “real time.” That is, you subjectively feel the
curring at the same time that your finger makes conta
table. But our experimental evidence strongly support
ing finding that is directly counter to our own intuitio
ings: The brain needs a relatively long period of appro
vations, up to about half a second, to elicit awaren
event! Your conscious experience or awareness of y
touching the table thus appears only after the brain
have become adequate to produce the awareness.
We are talking here about actual awareness of a sig
must be clearly distinguished from the detection of a
example, human and nonhuman beings can discrim
tween two different frequencies of tactile vibration, ev
the intervals between two pulses in each vibration freq
only a few milliseconds (msec) in length. A leading n
tist criticized our discovery of an interval of up to 50
fore a conscious experience appeared, on just these g
we can differentiate between vibrating frequencies in
cessive pulses are separated by a few milliseconds, how
when is one aware of that detection? Becoming
aware of the difference is what requires the relatively
In other words, detection leading to some response
unconsciously, without any awareness of the signal.
If such physiological delays are built into the brai
tion of sensory awareness, a number of profound qu
implications arise: Why do we feel as though we a
ately aware of an event, as if there were no delay in
awareness? What about our abilities to react to a sen
lus within 100 msec or so, a delay much shorter than
awareness? For example, is a competitive runner aw
sound of the starting gun when she takes off in a r
much less than 0.5 sec? Do unconscious mental funct
different time requirement than conscious mental fun
To be convinced of this unexpected and counterint
in awareness, you need to see the evidence. The fol
tions outline the kinds of observations we made and
led to the surprising discovery of a delay in awarenes
Initial Evidence from Cerebral Stimulation

In 1957 or so, my collaborator and friend, neuros
Bertram Feinstein, invited me to design and carry
ments that could be done while he performed sur
ments on the brain, and to do so in a way that intr
new risks and was acceptable to the patient. I jum
wonderful opportunity to investigate what the brain
order to produce a conscious experience.
34 •
T H E D E L AY I N O U R C O N S C I O U S S E N S O R Y A W A R E N E S S
conscious experience? What significant question cou
that was also amenable to experimental study, especi
the limitations of access time to the subject and to
structures available for study?
At the start, we had available electrode contacts sitt
surface of the primary somatosensory cortex (Fig. 2
the area of cerebral cortex that receives the direct sen
from all areas of the body and skin. It was also known
trical stimulation applied to the surface of this area c
subject who was awake, elicit a conscious sensation o
tingling or other responses. These sensations were re
the subjects as coming from some specific skin or b
ture, not from the brain. That is, the sensation is “refe
jectively to some bodily structure that normally sen
sory input to the spot of the cortex being stimulated.
Fortunately, we began with a relatively simple quest
led to some significant answers. The initial experime
tion became: What kinds of activations of neurons i
sory area are critical to production of just threshold
sensation, that is, the weakest reportable conscious
The relevant neuronal activations could be evaluated
effective electrical stimulations, and from the recorda
cal changes produced by the nerve cells.
The big advantage in studying this question by stim
brain itself was that we might find requirements at th
level that are obscured by stimulation of the skin. It w
known that a sensory input from the skin can result in
that ascend to the brain in several different spinal cord
T H E D E L AY I N O U R C O N S C I O U S S E N S O R Y A W A
[To view this image, refer to
the print version of this title.]

Fig. 2.1. Maps of human cerebral cortex.

A. Lateral view of left hemisphere. Rolandic fissure separates t
and parietal lobes. The anterior, frontal side of Rolandic fissur
the primary motor area (MI); nerve cells in this area send mo
fibers directly to the final motor-neurones that directly activa
muscles. The posterior area behind the Rolandic fissure contain
mary somatosensory area (SI). Nerve cells in this area receive
sensory nerve fibers that originate in the skin, tendons, and mus
Primary receiving area for auditory input is at the upper mar
temporal lobe. Primary visual area is at the rear (posterior) tip
cipital lobe.
B. Medial side of the left hemisphere. This side is at midline,
medial side of the right hemisphere. The frontal region is to

here, turned 180o from that in Fig. 2.1A.
The top end of the Rolandic fissure is visible, as it continues a
into the medial side.
Frontal or anterior to the end of MI is the supplementary m
(SMA). Electrical stimulation of the SMA can produce gener
movements or vocalizations. The SMA appears to be involved in
aration and initiation of a voluntary act (see Chapter 4).
The corpus callosum is the massive bridge of nerve fibers t
messages between the two hemispheres. It has been cut throug
order to separate the two hemispheres for this medial view.
The calcarine fissure, at the occipital pole (on the left here), m
location of most of the primary visual receiving area.
C. The sensory “homunculus.” Representation of the right side
in the primary somatosensory area of the left cerebral cortex. The
is laid upon a cross-section of the hemisphere at the level of S
cortex. From Penfield and Rasmussen, 1950.
For both MI and SI, the opposite side of the body is represe
right side for this left hemisphere). Also, the body is represente
down. That is, the head and face are at the bottom, and the legs a
the top of each area. From Penfield and Rasmussen, 1950. Reprin
permission from The Gale Group.
bral delay for awareness if we had been limited to
sensory input, from the skin in this case.
Another important experimental strategy was to
studying changes at the level for producing a just thr
sory experience. That is, we looked for the differe
brain activities between two conditions: 1) when the s
put was still too low to produce any sensory awaren
when the input was raised to a level that just begins
weakest reportable subjective sensation. This strateg
important benefits. First, it was clear that a normall
ing brain is necessary before any special neuronal ac
lead to a subjective sensation. With our approach,
having to deal with the enormously complex nece
ground of brain activity. Instead, we focused on t
events critical to the appearance of awareness, startin
general necessary background. Second, study of the
from no-awareness to the awareness of a sensory stim
give potential insights into what cerebral activities m
unconscious or nonconscious mental functions. (This
oped into an experimental study of the different re
for unconscious versus conscious mental functions.)
So, what did we find, from applying all kinds of va
the stimulus to the sensory cortex? (See Libet et al.,
1973.) The stimulus consisted of brief pulses of cu
about 0.1 to 0.5 msec in duration, in different experi
peated at 20 or 60 pulses per second. A time factor
to be the most interesting requirement for eliciting a
sensation. To elicit a report of a weak, threshold-l
38 •
How was this measured? With a long 5-sec train
pulses, the intensity (strength of current in each pulse
raised to some minimum (liminal) level in order to pr
weakest conscious sensation (see Fig. 2.2A). When t
intensity train of pulses was shortened below 5 sec, th
of the conscious sensation, as reported by the subjec
shortened. But the perceived strength of the sensatio
changed. Finally, when the liminal stimulus train was
to below 0.5 sec, the sensation vanished. Short trains
0.5 sec) could, however elicit a conscious sensation if
sity (peak current) of the pulses was raised sufficientl
2.2B). But the higher intensities got into a range that i
not often encountered in a person’s normal everyday l
ripheral sensory input.
How does raising the stimulus intensity make it p
trains of pulses shorter than 0.5 sec to become e
higher intensity would undoubtedly excite a greater n
nerve fibers and affect a larger number of nerve cells t
input from these fibers. Alternatively, that raising o
could result in an increase in the frequency of firing b
the same neurons that responded to the lower, limina
of stimulus. In this connection, a higher frequency o
pulses—for example, changing from 30 pulses per se
to 60 pps—resulted in a lowering of the liminal int
there was no change in the minimum train duration of
quired by the 60 pps to elicit a conscious sensation
2.2B). That indicates that the minimum requirement o
train duration is independent of the frequency or the n
Fig. 2.2. Duration of stimulus train of pulses, relative to prod

conscious somatic sensation.
A. Diagram of a train of 0.5-msec electrical pulses, at the lim
sity for eliciting any sensation, applied to cortical area SI, p
gyrus. From Libet, 1966.
Second line plots the amplitudes of the direct cortical respons
recordable with each pulse.
Third line indicates that no reportable conscious sensation is e
til the initial 0.5 sec of pulses has been delivered. The weak sen
gun after the 0.5-sec time continues at the same subjective inten
the stimulus train continues. (This is in contrast to stimulation o
tor cortex, MI; a motor response starts well before a 0.5-sec du
stimulus, and it builds up in strength as the stimulus continues.)
B. Stimulus train durations at intensities required for a thresh
tion to appear. (Diagram projected from data for many subje
there is a minimum train duration of around 0.5 sec (the utiliz
required for the minimally effective intensity to elicit a sensation
gle pulse usually elicited a motor twitch in the related body par
hand or forearm). From Libet, 1973.
40 •
stimulus pulses, if the liminal intensity for a given fr

used.
Raising the stimulus intensity could introduce a co
factor, in that different nerve fibers of smaller diame
be fired. How that might affect the responses of the
neurons is not clear and is difficult to deal with.
A further complication appeared with stimuli to th
sensory cortex of sufficient intensity to produce a resp
only a few or even a single pulse. However, these res
cluded a slight twitch in a muscle of the hand or a
these high intensities there was an observable motor
What the patients reported was then clearly related to
cular twitch, which generated an actual peripheral sen
sage from receptors in or around the muscle. These
sponses made it impossible to tell whether a single
sensory cortex is different from the one we obtained
stimulating the primary motor cortex (located in fr
sensory area). A few strong pulses to the sensory
duced repetitive slight twitches. The same stimuli d
the primary motor cortex produced a smooth contr
twitches), and this contraction could rapidly rise in st
eventually into a seizure with continued repetition
Clearly, the motor responses with stimulation of sen
were not due to electrical spread to the adjacent mot
We were able to settle the question of whether a st
stimulus pulse could elicit a conscious sensation wh
an electrode contact located in the ascending senso
below the cerebral cortex. A strong localized pulse h
elicit any motor response, and a 0.5-sec train of mu
pulses did elicit a sensation. In other words, some sub
ration of repetitive pulses is necessary to produce a
sensation; a single pulse is completely ineffective f
matter how strong (when no muscle twitch is produc
The requirement of some repetition of stimulus p
somatosensory cortex for production of a consciou
has been confirmed by several other groups (Gross
Tasker, personal communication; Amassian et al., 199
quantitative study established the minimum duration
tive liminal intensity pulses at the surprisingly larg
about 0.5 sec. A recent quantitative study of this requ
Meador and his associates (Ray et al., 1998, 1999) con
value in principle. However, in their case, the minim
tion required at the lowest effective intensity was
42 •
mal subjects and in the patients we studied.
Is this 0.5-sec requirement unique to the abnorma
activation, following stimulation at the surface of se
tex? The answer was no. The nerve fibers that carry
for sensations from the skin, joints, and muscles (with
tion of pain and temperature) proceed up to the brain
bundle on the back side of the spinal cord (see Fig.
terminate on groups (nuclei) of nerve cells in the
portion of the brain, the medulla oblongata. These ne
the medulla produce nerve fibers that cross over to
side of the brain and proceed to the forebrain in a bu
the medial lemniscus (so termed for its location and sh
crossover is what accounts for the representation of
in the side of the cerebral hemisphere opposite to the
origin of the sensory stimulus. (Thus, a stroke that da
pathway in the left side of the brain results in loss of
on the right side of the body. The evolutionary val
crossover is not clear).
The medial lemniscus fibers terminate on specific
nerve cells in a lower part of the base of the fore
thalamus. These “ventrobasal” cells in the thalamus s
fibers directly to the primary somatosensory cereb
This somatosensory area is located in a fold, or gyr
back of, or posterior to, the main vertical cerebral gro
the central fissure of Rolando. The spatial origins of
sensations are maintained in this whole pathway, and t
nate differentially on cells in a specific array for th
body parts (see Fig. 2.1C). The whole sensory pathwa
Fig. 2.3. Pathways for the touch, pressure, position-sense sens

fibers from the skin, tendons, and muscles. The fastest of th
fibers enter the spinal cord and send a branch directly up in the
or dorsal columns of nerve fibers. They end on cells in the me
lowest part of the brain. The nerve fibers from these cells cros
proceed up in a bundle called the medial lemniscus. These lemn
end on cells in the ventral-basal thalamus, which then send fib
on cells in the somatosensory cortex (SI, in the postcentral gyru
tical fold behind the central Rolandic fissure). The thalamus
base of the cerebral hemispheres and has other crucial functio
Chusid and McDonald, 1958. With permission from the Mc
Companies.
44 •
motor area, lying in front of the same central fissure.)
is represented in an upside-down way, with the legs a
the part of the gyrus near the top of the head, and th
head at its lowest end. So the representation is both cr
and upside down!
We had some cases in which electrode contacts we
in the thalamic parts of this system and in the medial
leading into the thalamus. These occurred when the
were placed in these structures for therapeutic purpos
duce a conscious sensation with electrical stimulation
of these locations, we found time requirements were
as they were for the sensory cortex. That is, the train o
the minimally effective intensity had to persist for abo
So, activations in this normal pathway to the cerebral
exhibited the same requirement of a surprisingly lon
of repetitive inputs to elicit a sensory awareness.
This new discovery of the temporal requirement fo
tivations at cerebral levels of the sensory pathway
seemed not to fit with that for stimuli at the skin, o
fibers from skin into the spinal cord. It has long be
that a conscious sensation can be produced even b
weak electrical pulse to the skin (or to the nerve fi
the skin). So, what goes on here? Is our proposal for
tial delay in awareness not relevant for normal input
skin?
To look at this question, we had to distinguish be
requirement for a peripheral (skin) input and that fo
bral processes to which this skin input gives rise. That
to test whether this statement is true: Is there a 0.5-se
a conscious sensory awareness even when it is gene
single pulse input to the skin?
Actual Delay in Awareness with Normal Senso

Even a single weak pulse to the skin or its sensory n
ficient to elicit a conscious sensation. This statement s
counter to the evidence cited in the previous secti
study we found that up to about 0.5 sec of activati
quired to give rise to a conscious sensation. If that ap
skin stimulus, a single effective stimulus pulse may h
duce a lengthy (0.5-sec) period of cerebral activations
conscious skin sensation can appear.
So, the next question was: Does the single skin pu
cerebral activations that must persist for about 0.5 sec
pulse elicits that conscious sensation? That is, is there
tual delay for sensory awareness when the message
as a single weak pulse applied to normal sources a
This question could only be answered by our disting
tween the input that is effective at the periphery (sk
activations set up by this input at the cerebral level,
lengthy requirement for awareness is in force. Indeed
not have discovered the time factor for awareness
been restricted to studying peripheral skin input, rath
rect intracranial input. We were, in fact, able to answe
tion we raised in the affirmative, based on three differ
evidence.
46 •
The first line of evidence deals with the electrical re
the cerebral cortex to the single effective stimulus pu
skin. It had already been demonstrated that each s
pulse gives rise to a sequence of cortical electrica
called the evoked potentials (EPs) or the event-related
(ERPs). These ERPs have been shown to represent ne
sponses in the cortex. They contain a number of diffe
nificant components (see Fig. 2.4). They begin with
EP locally produced in the sensory cortex in the spe
area to which the stimulated skin area “projects.” Th
the primary EP arrives via the fast specific projectio
discussed earlier. The primary EP begins after a del
some tens of milliseconds after the skin pulse. With
path, say from the hand, it starts after 14–20 msec, whi
path from the foot may take about 40–50 msec. The s
plitude of the primary EP is related to the strength of
from the skin.
A striking feature of the primary EP is that it is neit
sary nor sufficient for eliciting a conscious sensation.
that it was not necessary because we could elicit a con
sation with a weak stimulus applied to the surface o
sory cortex. This cortical stimulus does not produce a
electrical response equivalent to the primary EP; th
produced only by input arriving at the cortex from
the sensory pathway.
On the other hand, a single stimulus pulse in any p
specific sensory pathway that is located in the brain do
primary EP response of the sensory cortex. But this s
Fig. 2.4. Electrical responses (event-related-potentials, ERP) of

bral cortex to single stimulus pulses at the skin, averaged from
lus presentations at 1 per sec.
With just threshold strength stimuli (T) to the hand, virtua
components of the ERP are already visible. (At T strength not
stimuli were felt by the subject.) The initial uptick records the ti
stimulus pulse. About 30 msec later the first response at the SI
downward, surface-positive deflection, the primary EP. That is fo
later slower components, more pronounced with stimuli at twi
old strength (2T).
But note that stimulus with subthreshold strength (subT), at
of T elicits only the primary EP but no later components. (Ea
tracing is 125 msec long in A1 and B1, and 500 msec long in A
From Libet et al., 1967. Reprinted with permission from the Am
sociation for the Advancement of Science.
does not elicit any subjective sensation at all. This i

when the pulse is relatively strong and the primary E
it evokes is large (Libet et al., 1967; see Fig. 2.5). An
(single) responses from the primary sensory pathwa
a conscious sensation had also been observed by
Bertrand (1966). As already described, stimulus puls
48 •
Fig. 2.5. Single primary EP responses do not elicit any sensation. A
stimulus pulses to the ascending sensory pathway in the thala
VPL nucleus) does elicit a sensation, just as a train (0.5-sec durat
at the SI cortex. But single pulses to VPL, even at six times th
(threshold) intensity, which is effective with a 0.5-sec train of pu
per sec, elicit no sensation at all.
The initial 125 msec of the ERP response of the SI cortex is s
single stimuli applied to VPL and those applied to the skin (S). T
primary EP response to a VPL stimulus, at six times the liminal
for a 0.5-sec train of pulses to produce a threshold sensation, is la
the initial primary EP cortical response to single-pulse skin stimu
times the threshold strength of single S stimuli). Yet these single V
uli elicit no sensation, while these single S stimuli do elicit a m
strong sensation. The appropriate later components of ERP pro
the single S pulse (not shown in this 125-msec tracing) are not p
the ERP produced by the single VPL stimuli.
Note the delay (after the stimulus artifact) for the primary EP
shorter for the single VPL stimuli than for the S stimuli. That is
the VPL site is much closer to the SI cortex than is the S stim
hand. From Libet et al., 1967.
applied repetitively here to produce a conscious sens

as for stimuli at the sensory cortex.
Because the early primary response of the cortex
pulse) does not elicit a sensory awareness, some late
components must be required to achieve the awarene
under general anesthesia, the later ERP components
while the primary EP may even be enlarged; but of
patient does not feel any sensation. Similarly, if the
the single skin pulse is lowered to a level at which
normal subject reports feeling nothing, the late ER
nents are suddenly absent, but a distinct primary E
can still be recorded at the sensory cortex (Libet at al
It follows then, that the later responses of the cere
produced after a single pulse to the skin, appear to be ne
producing a conscious sensation. These late response
for more than 0.5 sec—long enough to provide the pe
tivations needed for the postulated delay in awarene
occurs even for a normal sensory stimulus at the skin
the actual minimum duration of these later evoked c
that are required for conscious sensation has not b
lished. Nor have possible specific components of t
sponses been identified as the specific agents for awar
Retroactive, Backward Effects

of a Delayed Second Stimulus
The second line of evidence is based on retroactive
effects of a delayed second stimulus, one that follow
testing one. Retroactive or backward masking betwe
ripheral sensory stimuli has long been known. Wi
stimulus consisting of a small weak spot of light, a se
ger larger flash that surrounds the first one can blo
jects’ awareness of the first one. The second flash ha
50 •
ulation of the skin (Halliday and Mingay, 1961). With a
ulus at threshold strength on one forearm, a suprathre
ditioning stimulus on the other forearm raised the thr
the test stimulus. The conditioning stimulus was effe
when it followed the test stimulus by 100 msec, but n
followed by 500 msec. This retroactive masking at th
interval must be mediated in the central nervous syste
the test and conditioning stimuli were delivered via dif
sory pathways.
What has this backward masking to do with our
delay in sensory awareness? If appropriate neural
must go on in the brain for up to 0.5 sec, to produ
ness, then a second stimulus delivered during that re
terval may interfere with the proper completion of th
tions and thereby block that sensory awareness. We
establish that such masking takes place in the respon
tures at the level of the brain, rather than in a peripher
structure. We also wanted to see whether the time in
tween the two stimuli, to produce the retroactive ef
be increased to something closer to our 0.5-sec require
To achieve these goals, we applied the delayed co
stimulus directly to the somatosensory cortex (see
The first (test) stimulus was a weak single pulse to the
delayed cortical stimulus was applied with a large 1-cm
trode. It was relatively strong and produced a sensatio
felt in a skin area overlapping the area of sensation pr
the skin pulse. The subjects had no difficulty in dist
A

Fig. 2.6. Retroactive effects of a delayed cortical stimulus that

single pulse stimulus to the skin.
A. Retroactive masking of skin sensation. A brief train of
pulses is delivered to the primary somatosensory cortex (SI), beg
msec (or more) after the weak stimulus pulse (at threshold—T—
to the skin. The cortical stimulus electrode is a 1-cm disc, appl
cortical area that “represents” the skin area being tested.
B. Retroactive enhancement of the subjective sensation, elicit
gle pulse stimuli to the skin. The cortical stimulus electrode is a 1
tact by a wire.
Top line: Two identical single pulse stimuli to the skin (S1 a
separated by 5 sec. Lower line: A train of stimulus pulses to S
cortex is begun at variable times following S2. After each trial, t
reports whether S2 felt the same as, weaker than, or stronger tha
Libet et al., 1992. Reprinted with permission from Elsevier.
mask or block awareness of the skin pulse even when
cal stimulus began up to 200 to 500 msec after the skin
cidentally, the delayed cortical stimulus consisted of
pulses. Cortical trains lasting less than 100 msec, or sin
were not effective for this retroactive inhibition.
We also made a surprising discovery that a delaye
could retroactively enhance, or intensify, the initial s
tion, instead of masking it. This occurred when we us
smaller electrode contact on the sensory cortex to pr
delayed stimulus. For this experiment, the initial weak
was delivered twice, the two equal pulses separated by
terval (see Fig. 2.6B). The subject was asked to repor
the second of these skin stimuli (S2) felt stronger, the s
weaker that the first one (S1). The cortical stimulus w
by intervals between 50 msec and 1,000 msec followin
ond, S2, skin pulse. The subjects reported that S2 fe
then S1, in most trials, when the cortical stimulus bega
to 400 msec or more after S2.
We subsequently found that a retroactive facilitati
hancement) had been reported by Pieron and Segal (1
the test and conditioning stimuli were both applied vi
electrode on the skin of a finger. The effect was seen
first or test stimulus was subthreshold. It became p
when the suprathreshold conditioning stimulus follow
stimulus by 20 to 400 msec.
Clearly, then, the conscious sensation elicited by a
pulse could be retroactively modified by a second inpu
delayed by about 500 msec. This adequately supporte
tant theoretical element in this support. For retroac
ing/inhibition, some have argued that the delayed co
ulus may simply disrupt the formation of the mem
skin stimulus that preceded it. That argument was in
on the fact that a generalized strong electrical stimulu
area of the brain (as in electroshock therapy) destroy
cent memories. But in such electroconvulsive shock t
plied therapeutically to patients with intractable de
large portion of the brain is excited strongly, thereby
a seizure. For our retroactive effects, the delayed stim
sensory cortex was both localized to a small area an
below the strength required to elicit even a local sei
cortex. The argument for a memory disruption in
masking is therefore very weak. But, with retroacti
ment, there is no memory loss at all. The subject rem
second skin stimulus as being stronger than the fi
stimulus.
Efforts to Deliberately Slow Responses

The third line of evidence appeared fortuitously in un
periments by Arthur Jensen (1979), a professor of psy
the University of California–Berkeley campus. Jensen
suring the reaction times (RT) of different groups of
these routine tests, subjects were asked to press a
quickly as possible after the appearance of an agree
nal. With the kinds of signal employed, Jensen’s su
duced RTs in the 200–300 msec range. Because there
ences in the average RTs between different groups o
Jensen wanted to rule out the possibility that some
54 •
msec or so. To his surprise, he found that none of th
could do that. Instead, they produced RTs of 600–
much longer than the requested smaller increases.
When Jensen heard about our evidence for a delay
500 msec for conscious sensory awareness, he rea
might explain his strange finding. To lengthen an RT
erate process, you may assume that the subject mu
come aware of the stimulus. Awareness of the stimulu
bly not required at the moment when the subject re
usual RT test, in which deliberating about the respo
an issue. (Indeed, there is direct evidence that ordina
accomplished before or with no awareness of the
But, to achieve awareness before a deliberate slowing
sponse, a requirement of about 500 msec of activities t
awareness would delay the response by that additi
That would explain the discontinuous jump in RT b
tional 300–600 msec when deliberate slowing of the r
attempted. This is the only available explanation of Jen
ing, and it provides additional convincing evidence fo
delay in sensory awareness.
How Does the 0.5-Sec

Neuronal Activity Lead to Awareness?
Is there a unique feature in the brain process that ex
a 0.5-sec duration of cerebral activations is require
awareness of an event? Are there testable options f
event? There are several possibilities.
First, this time requirement is unique for awareness
ness of the stimulus to that correct detection, we had
the duration of repetitive activations of sensory corte
0.4 sec. Clearly, awareness itself is a mental phenom
rate from the content of a mental event. Content o
can be detected by the brain unconsciously, without
of it.
Are special neurons fired after sufficient repetiti
Maybe the repeated neuronal activations produce a
rise in the excitatory level in some key neurons, so
neurons finally achieve a firing level. It would then
charge of impulses in such special nerve cells that lea
pearance of awareness. There is some evidence rela
opinion.
Stimuli to the sensory cortex, or to the ascendi
pathway in the brain, produce no sensory awareness
intensity of the stimulus pulses remains below a lim
lute threshold) level. (This level is that required to p
weakest sensation.) This is true even if the sublimina
repeated for 5 sec or longer. These subliminal puls
electrical responses of the cortex similar to, but sm
those for the effective liminal intensity stimuli. On
hand, it is possible that the subliminal intensity is
enough to ever excite some crucial nerve cell eleme
repetitive activation leads to the adequate excitatory
key neurons for awareness.
With a stimulus in the ascending sensory pathw
lemniscus), a single stimulus pulse can be made for
strong as each of the pulses in a 0.5-sec train of ten
56 •
level. That single pulse contains four times the elect
that the 0.5-sec train delivers at the liminal intensity
sation. This tends to contradict the idea that a simp
tive mechanism develops during the 0.5-sec train of
reaches an effective level for awareness. The strong si
might be expected to excite all the neurons that would
excited by the cumulative integrative effects of the 0
of weaker pulses.
A final point can be made against the idea that aw
sults from a special firing of some key neurons, at th
minimum stimulus train of pulses. This is seen in the
electrical responses at the sensory cortex (Fig. 2.7).
sponses exhibit no special change at the end, or jus
end, of the required 0.5-sec train. The responses are
tially identical during that stimulus train. However, on
mit that some special response could have occurre
nerve cells that was not recordable in our studies.
Robert Doty (my friend and an eminent neurosci
asked, “Is there something about the frequency of r
rather than duration per se, “that is the essential factor
attaining conscious experience depend on “the num
pulses being generated by the train” of stimulus pulse
sults with different stimulus frequencies do not sup
alternative suggestions. Stimulus trains to cortex at
quired a lower liminal intensity than those at 30 pp
minimum required train duration was similar for bot
frequencies, at their respective liminal intensities. Ther
ther the higher frequency nor the number of pulses w
Fig. 2.7. Direct cortical responses (DCRs) evoked in SI cortex b

stimulus pulses, at threshold (liminal) strength for conscious sen
Upper row: Single responses at different strengths of pulses: a
b at 0.8 mA (equal to liminal intensity, for 0.5 sec train of 20 puls
to elicit a sensation); c at 1.7 mA; d at 5 mA. Subject reported n
any of these single pulses. Horizontal bar in d is for 100 msec; v
is 200 µV.
Lower row e: 0.5-sec train of responses to 20 per sec pulses
(same strength as in b in the upper row, but at higher amplificati
Libet, 1973.
pps was significant. It was the duration of train that

to elicit awareness.
I have proposed a completely different option for
the 0.5-sec activity requirement for awareness: The d
similar activations may itself be the basis. That is, wh
58 •
for the emergence of awareness. This option fits w
presently available evidence. It is, therefore, a viable
though it cannot be said to be an adequately proven m
Role of Memory Formation

There is still another major issue in this question o
counts for the 0.5-sec duration of activations for aware
is the possible role of memory formation.
We have already noted that the only valid evidence
tive awareness is an introspective report of awareness
dividual who has the experience of it. Obviously, so
term memory formation must occur for the subjec
that awareness and report it. Incidentally, short-term
ing” memory is what accounts for our ability to recal
information some minutes following the event. The ab
call a 7- or 11-digit telephone number, after seeing it
example of this type of memory. Without further
person would be likely to forget the number within
Long-term memory involves an additional neuronal p
lowing its effect to persist for days, months, or years.
Some academicians have argued that the 0.5-sec d
activities required for awareness is simply a reflecti
time it takes to produce the short-term memory tr
event (see Dennett discussion in Libet, 1993). There a
two ways in which this memory formulation might o
one case, the production of the memory trace woul
regarded as the “code” for awareness. In the other ca
ness of an event would appear without any signific
which I will discuss briefly.
Explicit Memory and the Hippocampus Structu

Reported observations of human subjects provide a m
ment against a role for memory formation in the
of awareness. In both human and nonhuman anim
structures within the temporal lobes of the cere
spheres are necessary as mediators for the formation
declarative or explicit memories. These kinds of me
those that can be consciously recalled and reported
distinguished from nondeclarative or implicit memor
memories are formed without any conscious aware
event, and they cannot be consciously recalled and
They function largely in the learned acquisition of
mechanical and intellectual.
The hippocampus structure in the temporal lobe
ronal component that is necessary for mediating the
of explicit memory. Even if one hippocampus were
the intact structure on the other side of the brain c
out the memory process. But if both hippocampal
were destroyed, the person would suffer a profound
ability to form new explicit memories. Such an indivi
have virtually no recallable awareness of events th
taken place. He could not tell you about an event ev
ately after it occurred.
Such losses have resulted due to a pathological les
temporal lobes. More definitively, this bilateral loss
when a surgical procedure to remove an epileptic fo
60 •
good structure was excised, leaving the ineffective pa
structure on the other side. This mistake led to the di
the role of the hippocampus structures in explicit me
mation.
Now, the interesting observation, for our present p
the following: An individual with a bilateral loss of
campal structures has virtually no recallable awaren
event or sensory image that has just happened (altho
term memories formed before the loss are recallable).
such an individual retains the ability to be aware, both
mediate present and of himself.
A movie of a patient with this type of loss shows a
is alert and communicative. He is clearly aware of his
ings and of the psychologist who is interviewing him.
aware of his own inability to remember what just
and he complains about that as a significant loss in the
his life.
This patient had in fact not lost all memory form
was able to sit at a computer and learn to play a gam
He was unable to explain how he acquired the skill. T
ries for the learned skill were obviously of the imp
which do not require the function of the hippocam
tures: A different neural pathway must be involved. B
no awareness associated with implicit memory. And s
memory cannot be employed as an argument that m
a role for producing awareness.
There is some question about whether there is form
declarative memory that persists for at least 0.5 se
who studied the patient just described were confiden
patient could remember things for a minute or so.” O
hand, studies of similar patients employed psycholo
tive tests that did not demand evidence of a consciou
example, Drachman and Arbit, 1966). It is possible
that the observed short-term memory was actually e
some nondeclarative, implicit memory. In that case, it
be relevant to the question of the role of memory i
of a conscious experience. In any case, Larry Squire
researcher in the field of memory processes, has ex
opinion that conscious experience is independent of
of memory formation (personal communication). It
pear, then, that the retention of awareness in peop
severely reduced ability to form new explicit mem
cates that the phenomenon of awareness is not a f
a memory process. That fundamental observation
any hypothesis that awareness depends on a form
memory.
Classical Conditioning and Awareness

Clark and Squire (1998) discovered an interesting role
ness in classical conditioning. In classical conditionin
tioning stimulus (CS) is presented just before and du
(unconditional stimulus). The CS can be a tone tha
produce the response initially; the US can be an air
eye that elicits an eye-blink response. After some tr
combination, the subject (human or experimental
62 •
with bilateral hippocampal lesions. In trace conditionin
arranged to end about 500 to 1,000 msec before the on
US. Animals with bilateral hippocampal lesions fail
trace conditioning. Amnesic patients with damage to
pal formations are able to learn standard delay conditi
also fail to learn and perform trace conditioning—
the experimental animals. Normal human voluntee
course, acquire the trace conditioning, but only whe
aware of the stimuli. So, trace conditioning not only d
the hippocampal structures, but it is also somehow c
the process for awareness.
Now, these findings do not prove that production of
tive memory is the basis for the roughly 0.5-sec duratio
bral activities needed to elicit awareness of an event.
Squire (1998) suggest that
[a] conjoint operation of the hippocampal system

the neocortex may be a critical element that c
awareness about the (declarative) knowledge tha
been acquired. . . . But that does not mean that a
ness, per se, requires the memory function of the h
campus. Indeed, the existence of awareness in th
sence of declarative knowledge, in patients with bil
loss of hippocampal systems, supports the view th
mation of a declarative memory is a separate p
from the unique process for producing awareness
The finding that trace conditioning requires subje
tioning, the best studied of all learning paradigm
register with current understanding of the memo
tems of the brain.
An important implication of this finding is that trace

ing may offer a possible method for studying awaren
human animals. Simple delay conditioning is nondec
formation does not require the hippocampus or awa
exhibited by amnesic patients who lack short-term
memory.
Other Evidence on Proposals of Memory for Aw

Even though the foregoing evidence appears to rule
ory formation to explain the 0.5-sec activities required
ness, it is interesting and instructive to analyze at lea
proposal. After my lecture in a symposium on con
sponsored by the Ciba Foundation (in London), the p
Daniel Dennett proposed that the conscious aware
event might appear almost immediately, as it does in
to do for a stimulus at the skin. But, he argued, the
cannot be recalled and reported unless there is a suffi
tion of neuronal activity to produce and “fixate” th
for that awareness. Dennett’s argument was also inte
clude the need for postulating a backward referral of
tive timing of a sensory awareness, as discussed belo
cussion in Libet, 1993b, pp. 140 and onward). At the
not recall the evidence I have already cited: Declarat
memory is not necessary for awareness, and both m
64 •
As I have already noted in the second section of this
is possible to inhibit or mask the appearance of a con
sory experience, if a weak sensory stimulus is followed
of stimulus pulses applied to the sensory cortex. This
masking occurs even when that train does not begin
500 msec after the skin pulse. That result demonstrate
layed input can interfere with the content of the senso
ences. I cited that data as evidence that a neuronal a
riod is required to produce the sensory awareness.
Dennett countered that the delayed masking stim
simply disrupting the formation of the memory tra
awareness. (Electroconvulsive shock therapy is indeed
disrupt recent memory formations. However, the dela
ing stimulus employed in our experiments is minuscu
parison to the strong generalized electric shock used
therapy.) But his argument is countered by two o
rimental observations: (1) A second masking stimul
applied after the first masking stimulus (Dember an
1967). The second masker wiped out the sensations fo
masking stimulus, and the awareness for the original s
lus then reappeared. That means the first masking st
not wipe out the memory trace for the original skin st
When the delayed cortical stimulus is applied by a mu
electrode contact, the original skin pulse is not mask
tually felt to be stronger (Libet et al., 1992). With this
enhancement of the sensory awareness for a skin stim
was clearly no loss of memory at all.
So, the retroactive effects of a delayed stimulus on
for the original skin pulse (during the 0.5-sec delay).
Max Velmans also made an ingenious argume
Dennett’s proposal (see discussion in Libet, 1993b, p
As Velmans pointed out, Dennett’s proposal that sen
be experienced early but then forgotten is not exp
falsifiable. For example, in a standard psychophysical
you can establish the threshold for awareness of a sen
lus. As you gradually turn up the intensity of the stim
tain point is recorded at which the subject says she can
(or see it, or hear it). Then you can lower the intensi
subject says she does not feel it. The subject’s repo
related with the strength of the stimulus and are a
accurate and valid. But, according to Dennett’s pr
subject’s inability to report feeling the weaker, below
stimulus could result from rapid forgetting of an actu
enced event. “Dennett could extend that claim to a
that subjects make about not having experienced s
Dennett’s view, in other words, could never be con
he would not accept a subject’s report of no sensory
Such proposals are not scientifically acceptable becau
speculative, untestable beliefs.
I conclude, then, that awareness is a unique ph
with its separate neuronal requirements. Awarene
function of a memory process. It is not the equiv
formed, declarative memory trace. Nor is the absen
port of awareness due to a rapid forgetting of an e
sensory experience. The proposal that remains most
66 •
What Does the Primary Evoked
Cortical Response to a Sensory Stimulus Do?
You may ask, if the cortical activities that produce th
primary EP appear to have no essential role in beget
sory awareness, what function does the primary EP
primary neural response is important for the discrim
the precise location of a stimulus at the skin. And, as
ered, it appears to provide a timing signal to which t
subjective timing of the input to the skin is retroactivel
In some forms of a cerebral stroke, there is major d
this fast, specific sensory pathway as it approaches th
cortex. These stroke patients can locate a skin stimulu
very crude way, and they lose the ability to tell the
points of a two-pointed stimulus to a hand until thes
points are separated by many centimeters.
In addition to this spatial deficiency, we found, in th
patient available to us, that a touch pulse at the skin
ceived as delayed by about 0.5 sec, compared to a pu
to the normal side (see Libet et al., 1979). This p
a stroke some years earlier, confined to an area of
hemisphere. This stroke left her with permanent dam
specific ascending sensory pathway for bodily sensa
showed an inability to accurately localize a stimulus
hand or arm, and could only report its location very c
tested this patient for her subjective timings of stim
When stimuli were applied simultaneously to both
subject reported that she felt the right hand stimulus
(bad) left hand one. The stimulus to the affected sid
delivered about 0.5 sec prior to the stimulus to the n
in order for the patient to report that both stimuli we
sciously as being simultaneous. Clearly, she had lost
to subjectively refer her sensation in the left hand b
time. That sensation was therefore subjectively time
roughly 500-msec delay imposed by the cerebral requ
awareness. That loss in the ability to antedate the awa
presumably due to her loss of the primary evoked r
her left hand.
Conscious synchronicity of sensations. This brings up

tant general question about how different stimuli t
tually delivered synchronously can be consciously p
being synchronous. With stimuli in the same som
modality, there are different conduction times in t
pathways, depending on the different distances be
stimulus locations on the body. The time for the ar
fastest sensory messages varies between 5–10 msec (fo
the head) to 30–40 msec (for stimuli to the feet). Be
chronous stimuli to these two areas are subjectively
as synchronous, we can only assume that a time di
30 msec or so is not subjectively meaningful. On the o
a very strong stimulus to one site may require a s
shorter duration of appropriate cerebral activities; the
can be as much as 100 to 200 msec for the two differen
68 •
nificantly shorter cerebral activation times may not o
monly.
What about synchronously applied stimuli in differe
ties? Suppose that a flash of light appears at the same
crack of noise, both from the firing of a gun. Of co
travels faster than sound; but if the gun is fired just
away, the difference in travel times would not be sign
a speed of about 1,100 feet per second, the sound w
the listener in about 2 msec from 2 feet away). As wit
sensory stimuli to the body, the visual and the audito
also elicit a fast primary evoked potential in the visua
tory cortex respectively. The latency or delay for arr
fast signal to the visual cortex is distinctly longer th
other modalities. That is because the retina takes ext
get from the photoreceptors through the next neural
der to fire the ganglion cells, which send impulses vi
nerve fibers toward the thalamus and on to the visu
The delays for the visual primary evoked response in t
brain have been measured at about 30 to 40 msec by
(1977).
The primary evoked responses at all sensory cortex
calized to a small area that represents the peripher
spot or area being stimulated. With a recording electr
surface of the cortex, a substantial primary evoked po
only be recorded at the “hot spot,” the spot of cortex
the fast input from the peripheral sensory elements th
to the sensory stimulus. The primary evoked potentia
not significantly seen in recordings made with an ele
in the tissues between the cortex and the scalp. Co
the earliest significant electrical potential seen in a sc
ing is a later component of the response to a stimulus
latency of 50–100 msec longer than the time primary
tential, and it would be misleading to use the later
consideration of the problem of synchrony among
multaneous stimuli.
In any case, the true primary evoked potentials can
cies between about 5 and 40 msec, depending on t
and modality of the stimulus. If all simultaneous stim
ceived subjectively as being synchronous, we would
sume that the brain does not “consider” this range of
in latencies as subjectively significant.
Why Get Excited about the Delay in Sensory A

If we look at some of the ramifications of our findin
lay in sensory awareness, the implications are quite a
We consider a larger variety of important implicatio
this section, but I mention a few obvious ones here.
First, if awareness of all sensory stimuli is delaye
0.5 sec, following the pattern found for somatic sensa
our awareness of our sensory world is substantially delay
actual occurrence. What we become aware of has a
pened about 0.5 sec earlier. We are not conscious of
moment of the present. We are always a little late. I
how can one explain the fact that subjectively we fe
70 •
ject may be considerably different from the actual im
to the subject. For example, if a prudish man were sh
ture of a naked lady, he might report seeing somethin
ferent, or he might report that he saw no image. T
would not be consciously and deliberately distorting
instead, he would appear to believe that he was givin
of what he saw. That is, the distortion of the content
take place unconsciously. Sigmund Freud was, of cour
the first to call attention to what he postulated was
sion of conscious content. That suppression would b
about by unconscious processes that “protect” the su
an unpleasant conscious experience.
Given such unconscious modification of what w
aware, there must clearly be some delay in the awaren
which such a subjective modulation can be produced
ness of a sensory image were produced almost im
with no substantial delay, it would be very difficult t
how an unconscious cerebral process could then be m
produce a modified content of that awareness withou
ject knowing it.
Our discovery of the substantial cerebral delay for
thus provides a physiologically required time inter
which other inputs may modulate the content of an
before it consciously appears. As previously described
active effects of a delayed cortical stimulus can, in fac
sciously change the conscious content of a skin sensa
ported by the subjects in our experiments.
We would have to modify the existentialist view of l
experience of the “now”; our experience of the “now
delayed or late.
Further, there is the possibility provided for eac
character or past experience to alter the conscious
each event. That means each person has his or her o
ual conscious reality. The 0.5-sec delay for awareness
makes that possible. Differing perceptions of reality
meaning for the different paths that people follow, bas
individual’s conscious perceptions of reality.
In any case, our knowledge of a substantial delay
ness shakes up our confidence in our certainties abo
of the world.
Antedating of Delayed Sensory Experience

The evidence appears to show that some appropriat
activities in the brain must endure for up to about 50
even a single-pulse skin stimulus, in order to elicit a
sensory experience. But subjectively we seem to be
skin stimulus almost immediately, with no appreciab
we have a strange paradox: Neural activity requirem
brain indicate that the experience or awareness of a
lus cannot appear until after some 500 msec, yet subj
believe it was experienced without such a delay.
This troublesome dilemma bothered us for some t
began to think that subjective timing need not be i
neuronal time (in other words, the time when the n
72 •
Fig. 2.8. Subjective timing of sensory experience with cerebral ve

stimuli.
Cortical stimulus train at 60 pulses per sec, at liminal inte
threshold sensory experience. Because the train must last 500 m
sensation it produces cannot begin before the 500-msec time perio
gle threshold pulse to skin (S) is applied at the 200-msec duratio
cortical train. If that also required 500 msec to be developed, the
tion should be felt after the C-induced sensation. But subject repo
sensation felt before the C-induced sensation. That held true wh
pulse was further delayed. But when the S pulse was applied nea
msec time of the cortical train, the subject reported both sensati
felt at the same time. From Libet et al., 1979. Reprinted with pe
from Oxford University Press and Brain, vol. 102, pp. 191–222.
tually produced the experience). Indeed, we ran an e

that demonstrated this discrepancy directly (Fig. 2.8; L
1979). For this test, the train of stimulus pulses (near
strength for awareness) was applied to the sensory c
required about the usual 500 msec of repetition to
conscious sensory experience. (This cortically induced
was reported to appear in an area of skin like the ha
not felt to appear in the brain.) We then added a sin
threshold pulse to the skin. This pulse was applied a
times after the start of the cortical train, in different t
appeared before the cortically induced sensation, even
skin pulse was delayed by some hundreds of millise
the start of the cortical stimulus. It was only when the
was delayed by about 500 msec that the subjects repo
that both sensations appeared almost simultaneou
subjective time of the skin-induced experience appea
no delay, relative to that for the cortically induced
The cortically induced sensation was delayed by abou
relative to the skin-induced sensation.
We already had good evidence that awareness of a
does require about 500 msec of activities in the brain
that found with a cortical stimulus. Yet the skin puls
to be timed subjectively as if there were no such sub
lay. How could we deal with this paradoxical empirica
Is there a mechanism in the brain that could account
crepancy?
The clue to a possible answer came from the dif
tween the electrical response of the cortex to a sk
versus to a cortical-surface stimulus. The skin pul
characteristic response of the sensory cortex that be
wave or component about 10–30 msec after the ski
This is the primary EP, which is followed by later E
components. However, the stimulus pulses applied
face of the sensory cortex do not elicit any response
bles the primary EP (at least not with cortical stim
range of intensity we used).
This difference in cortical EP responses to the tw
sites of stimulus (skin versus cortical surface) led me
74 •
msec of appropriate brain activities. But then, there is
referral of the timing for that experience back to the time
mary EP response! The primary EP response of the cor
only about 10–30 msec after the skin stimulus, dep
how far the stimulated skin is from the brain. This dela
msec is not sufficient to be experienced consciously. T
ence or awareness of the skin pulse would thus be ant
ferred backward in time) subjectively to the timing
vided by the primary EP response. The skin-induced
appears subjectively as if there were no delay, even tho
not actually appear until after the 500 msec require
ronal adequacy to elicit that sensory experience.
This rather outrageous hypothesis could not be ser
posed without an experimental test of its validity. (
mental test, or at least its design, is mandatory for an
hypothesis.). Fortunately, we were able to devise an
crucially effective experimental test.
The test was based on the interesting fact that stim
specific ascending sensory pathway in the brain (in ot
in the medial lemniscus bundle; see Fig. 2.3), have tw
features. First, to elicit a conscious sensation, stimuli r
rations up to about 500 msec, just as for stimuli at th
cortex. Second, each individual stimulus pulse in a 500-
of pulses in the medial lemniscus elicits a fast prim
sponse recordable at the sensory cortex. This is the sa
response of the sensory cortex to a skin stimulus. It is
stimuli at the surface of the sensory cortex, which d
duce any such primary EP.
train in medial lemniscus should make the subjectiv
timing of sensory awareness the same as that for a
(Fig. 2.9). So in the experimental test, we matched
train of stimulus pulses to medial lemniscus with a s
tive pulse to skin. This experiment was similar in de
discussed earlier, when a skin pulse was matched wit
pulses delivered to the brain’s sensory cortex. The s
asked to report which of these two sensations appear
tively) first, in other words, the sensation elicited by
lemniscus versus the sensation elicited by the skin pu
The results of that experiment, to our delighted su
firmed the prediction from our hypothesis. When the
was delivered at the same time as the start of the m
niscus train, subjects tended to report that both sen
peared at the same time. But we knew that the sub
not possibly have experienced the medial lemniscu
until the required 500 msec (or 200 msec with strong
had gone by. They felt nothing if the train of pulse
lemniscus were cut to less than the required durat
msec. As in the case of medial lemniscus, the skin p
tion was reported to appear before that from the cor
lus. Only when the skin pulse was delayed until the
required cortical train were the two sensations repo
pear at the same time.
So, we see that even though both the cortical and
lemniscus stimuli required similar durations of repet
to produce the sensory experience, the subjective tim
76 •
Fig. 2.9. Diagram of hypothesis for subjective referral of a sensor

ence backward in time (antedating of experience).
The average evoked response (AER) was recorded at the SI cere
tex in response to single pulses, just above threshold for sensation
(at the S pulse arrow) to the skin of the hand centralateral to th
ing site.
The first dashed line below the AER shows the delay to
“neuronal adequacy” for producing the sensation (based on o
dence). The second dashed line shows the putative subjective re
the sensory experience back to the time of the initial primary ev
tential in the AER. This accounted for the observation that the s
ports the time of the experience as showing no significant delay
stimulus to the skin. From Libet et al., 1979. Reprinted with pe
from Oxford University Press and Brain, vol. 102, pp. 191–222.
stimulus elicited a primary EP response to each of i
shares that effect with the single pulse to the skin.
Some have raised a criticism that the test is based o
natural” nature of the medial lemniscus stimulatio
brush that aside by simply comparing the results whe
lemniscus stimulus is matched against a cortical stim
of these stimulus sites are “unnatural,” but any di
their behavior is clearly meaningful. Because the min
ronal delays for the experience were similar for both
earlier subjective timing for the medial lemniscus st
be taken as direct evidence for backward referral of it
timing of the sensory experience.
We have, then, produced strong and direct eviden
subjective timing of a sensory experience is antedate
later minimum time at which the brain activities a
come adequate to produce the experience. The sub
ing is referred backward to a “timing signal”: the prim
sponse of the sensory cortex. That accounts for our
feeling and belief that we become aware of a sensory
tually immediately, even though there is in fact a
delay!
This subjective referral backward in time, of our
sensory awareness, also accounts for another phenom
may ask, what happens to subjective timings of dif
sensory stimuli, which are different in strengths and
but are all delivered at the same time? Subjectively,
believe that all these stimuli were applied synchronou
78 •
Subjective Referral of Experiences
We have just seen, in the previous section, that the
timing of a sensory experience may not correspond w
layed time at which the cerebral neurons actually elic
perience. The sensory experience is automatically a
sciously subjectively referred backward to the time o
fast cortical response to the sensory signal.
An analogous situation was already known for the
age of a sensory event. The sensory image that one
aware of subjectively looks very different from the s
tern of neuron activities that elicited that image.
The most obvious and direct illustration of subjec
rals, being felt at different locations from the site of a
ulus, can be seen when you directly stimulate the cere
tosensory cortex. The subject does not feel or expe
resulting sensation as located in the brain, where it
duced. Rather, the subject feels the sensation to be lo
place in the body that is normally related to that s
brain. For example, if a certain cortical spot is stimu
subject feels that her hand has been stimulated. She s
refers the spatial location out from the brain to so
structure. She is completely unaware of the activity in
lated area of the sensory cortex.
Descartes theorized in the seventeenth century tha
rience of a sensation occurred only when the appro
of the brain was excited, but that the subjective locat
proposal is amazingly similar to the present direct ev
porting this view.
A related sort of subjective referral occurs in the s
tions for all normal sensory inputs. Take a visual i
light pattern from the visible object activates a patte
nerve impulses that reach the visual cortex after
through some relay stations. The spatial arrangem
cortical nerve cells that respond to the optical input a
very distorted pattern, compared to the actual image
to the eye. In fact, the original visual object would no
nizable from a map of the cortical responses to its lig
The cortical response may be greatly distorted spa
compared to the image that we subjectively see. But,
the spatially distorted cortical representation that g
the image you experience. We must therefore conclu
brain’s distorted neural pattern, in response to a visu
subjectively referred or projected into space, in such
the image seen corresponds better and more accura
actual visual object.
We have now established experimentally that there
jective referral of the timing of a sensory experience.
way analogous to spatial referral. Although stimula
cerebral sensory pathway to the cortex, may have to
up to 0.5 sec to elicit a conscious sensation, subjects
subjectively the sensation appears without any signifi
The delayed experience is subjectively referred back
the timing signal provided by the fast primary EP r
the sensory cortex. Subjects unconsciously and autom
80 •
stimulation of up to 0.5 sec in duration had taken plac
Subjective Referrals “Correct” the Neural Repre

Subjective referrals of the spatial and temporal feature
sory event have the effect of subjectively correcting th
distortions of the sensory event. The distortions are im
the way in which the cerebral neurons represent the e
in space and in time. So, in our conscious experience
sory event, the event seems to occur when it actually
instead of 0.5 sec. later (when we, in fact, became aw
event). It is of further interest that the specific projec
way to the sensory cortex provides the signal used b
spatial and the temporal referrals. Damage to this path
some cerebral strokes, results in losses both in subje
izations of a stimulus and in subjective timings of the
An unconscious ability to localize stimuli is availab
phenomenon of blindsight, subjects can localize targ
rectly pointing at an image, even though a lesion of th
visual area has destroyed the specific primary projecti
and the subjects do not consciously see the targets.
that the specific sensory projection to the primary vis
required only for subjective localization, not for unco
calization.
The subjective “corrections” of the information su
the sensory cortex are apparently learned. The most
dence that subjective referrals are learnable phenom
from some amazing experiments reported some year
man subjects were fitted with prism spectacles that t
field.
However, after wearing these spectacles contin
about a week, subjects began to be able to behave a
age were normal. A subject replying to a question abo
jective experience reported that he was not aware
verted visual image; but, when asked, he recalled it a
look upside down! Objectively, of course, the visua
still reversed from its normal arrangement. In other
subject somehow learned to not pay attention to th
and adapted his visuomotor responses as if the imag
right. When the spectacles were removed, the subj
motor accuracy was again briefly less accurate; it
within a few days. The experiment indicated that t
behavioral change was not due to an actual subjectiv
of the image; rather, the awareness of the upside-d
was somehow suppressed.
Interestingly, such flexibility of the visuomotor refe
anism is not seen in all animals. When Roger Sp
turned the eyes of frogs around, so the eyes we
down,” the frogs always reached into the upside-d
fields seen by those eyes. They did not learn to respon
verted image correctly.
These characteristics of visuomotor referral sugges
born infants may have difficulties with their still-not
organized visual images. They may need to learn to b
refer the visual information in a way that produce
that corresponds to the real sensory image. Perhaps
tially see a distorted image that corresponds to the dis
82 •
to see properly until about a month or so after birth. C
sion experts devise a way to test such a hypothesis?
Then there is the interesting question of how eac
sensory area of the cerebral cortex elicits its own spec
of a subjective sensation, when activated by a prope
stimulus or by the arrival of its normal sensory in
is, stimulation of the somatosensory cortex in the p
gyrus elicits subjective bodily sensations (touch, pre
tion, warm, cold, though not pain). Stimulation of
cortex (striate cortex in the occipital pole at the rear
sual sensations. Stimulation of the auditory cortex (u
the temporal lobe) elicits sensations of sounds. Altho
is some difference in the arrangement of the neuron
different areas, the basic neuronal structures and the
interconnections are similar. Stimulation of most oth
the cerebral cortex does not elicit any conscious e
That is not because the nerve cells are unresponsive t
tion in these nonsensory areas. It is presumably due
ability of the electrically excited nerve fibers to activa
complex network of nerve cells that serve the functi
nonsensory areas.
Well, each sensory input becomes able to produce t
subjective quality of sensation normally produced in
A question then arises as to whether the same specific
would be elicited if the sensory input to the area wer
That question has led some scientists to pose a strange
What would happen if we led the acoustic sensory pa
a functional connection with the visual area of corte
of course, be carried out on human subjects. But i
in a limited way, carried out in ferrets (Sharma et
Melchner et al., 2000).
In the neonatal ferret, the ascending retinal pathw
routed into the MGN (medial geniculate nucleus)
cleus normally receives the ascending auditory pa
then projects its axons to the auditory cortex in th
lobe.) The normally ascending auditory pathway to
was cut below the MGN. The animals were then rais
hood before testing. The researchers found neurons,
mally auditory cortex, that responded to visual stim
“rewired” neurons were organized into orientation
comparable to normally innervated visual cortex. I
the ferrets responded to light stimuli (presented in t
of visual field “seen” only by the altered projection)
they perceived the stimuli to be visual rather than a
the perceptual modality of a primary sensory cortex
by the input it receives. Of course, the ferrets could
researchers what they subjectively perceived.
General Implications of Subjective Referrals

Subjective referrals of cerebral sensory responses, sp
temporally, depend on the availability of the fast an
primary responses of sensory cortex. In the absence o
mary responses, the subjective referrals become ina
completely absent. But perhaps there is another sens
all subjective experiences are “referred.” Subjective e
of mental events, generally, are not reducible to or
by the activities of the nerve cells that give rise to
84 •
introspective report of the experience by the indivi
alone has “access” to his conscious functions. Put
way: The nature and content of a conscious experien
“look like” the pattern of nerve cell activities that pro
experience. You might therefore propose that the n
content of all subjective experiences are referred into
sphere, from the nerve cell patterns of activity that see
the experiences.
Neural Mechanisms for Subjective Referrals?

There is another aspect of subjective referrals that m
fundamental importance to the nature of the mind-
tionship. There appears to be no neural mechanism tha
viewed as directly mediating or accounting for the ref
Take the case of how the primary evoked potential
the cortical response to which the subjective spatial lo
the subjective timing of a sensory stimulus are refe
does that happen? This primary EP appears even when
stimulus is below the threshold for sensation. It the
alone, without any later evoked EPs. The later co
EPs lasting 0.5 sec or more, appear when the stimulu
equals and exceeds the threshold for sensation (see L
1967). The primary EP is exhibited only in a highly
small area of sensory cortex. But the later EPs are no
to the primary sensory cortex; related responses are b
tributed in the cortex. A broad spread of activities w
gle visual event, presumably one above threshold, ha
scribed by others (see Buchner et al., 1997).
It is the primary EP response alone that appears
referral of a delayed sensory experience back to the p
response, although such a mechanism is not imposs
primary EP response is the timing agent without the
by other unknown neural actions, it would appear t
tive referrals are purely a mental function, with no c
ing neural basis in the brain.
But the issue of mental functions related to neura
broader than the specific one of subjective referrals o
sensory experiences. All the brain processes that give
jective conscious experiences (including thoughts,
self-awareness, and so on) do not “look like” the eme
riences. Indeed, even a complete knowledge of the
neural processes would not, a priori, describe the acc
mental events. (The two phenomena have to be studi
to discover correlations.) The transformation from ne
terns to a subjective representation would appear to
the mental sphere that has emerged from that neuron
(The use of some specific neural signals to guide the
ferrals does not tell us how the referral is achieved.)
How does the inference of no direct neural descript
jective sensory referral and other mental events relat
philosophical views of the mind-brain relationship? F
proposal does not invoke or constitute an instance of
a Cartesian sense. That is, the proposal does not enta
ble or independent existence for the physical brain an
tal phenomena. My view of mental subjective functi
is an emergent property of appropriate brain func
conscious mental cannot exist without the brain pro
86 •
duced it. This view follows that espoused by Roger
emergent properties of a system (for more detail, see
and 6).
Identity theory is probably the most commonly h
sophical theory for relating the “physical” to the “m
Hook, 1960). In a simplified version of identity theory,
nally observable features of the brain’s structure and
in other words, the physically observable aspects—de
external or outer quality of the system. The mental e
scious or unconscious, describe an “inner quality” o
system or “substrate.” That is, the given substrate is r
for both the outer and inner qualities described. Iden
recognizes that subjective experiences are accessible o
inner quality) to the individual who has them. But, if
specific neural (physical event) that corresponds to
event (like subjective referrals in space and time), th
no common substrate to provide the identity for these
inner qualities. One of the early leaders for identity t
the late Stephen Pepper (1960), Professor of Philoso
University of California–Berkeley. In my discussions w
sor Pepper, he promptly realized that our findings of
referral backward in time would create serious diffi
identity theory. That is particularly true if there is
counterpart for this mental operation.
Identity theorists may say that this apparent disconn
tween observable and inner (mental) qualities is simp
the two aspects (the outer and inner) of their comm
substrate are expressed. But that would seem to glos
clear that inner, mental phenomena have features qu
from those of the physically observable brain, and tha
and outer qualities are each not describable, a priori,
the other.
A separate issue is how to regard one’s view or ex
the present, the “now.” The delay, of up to 0.5 sec, in
ance of awareness of a sensory event introduced a d
how to define or understand “the present moment.
existence of subjective referral backward in time (to
the fast primary response of sensory cortex) does pu
tive experience of the present back into the present. S
the strange situation in which actual awareness of
is really delayed, but the content of the conscious e
brought into alignment with the present. Subjectivel
do live in the antedated present, although in fact
aware of the present for up to 0.5 sec after the sensor
rives at the cerebral cortex.
These implications have serious consequences
views of the present. For example, Ludwig Wittgen
ported to have stated: “The present is neither past or
experience the present is therefore a phenomenon wi
ness.” But if our experience of a sensory stimulus is
tedated after the 0.5-sec delay, the experience is actu
an event 0.5 sec in the past. And so, the subjective “pre
tually of a sensory event in the past; it is not “timeles
Endogenous conscious events (our thoughts, im
nonsensory feelings, and so on) differ from normal
perience. Referral backward in time, the antedating o
88 •
scious events do not have such a timing signal availa
endogenous conscious events also require up to 0.5 sec
priate neural activations to produce awareness—in ot
if they follow our putative principle of this requirem
awareness—then the endogenous conscious events w
experienced after a delay. The delay would be from
ning of the unconscious neural events postulated to
awarenesses.
3
UNCONSCIOUS AND CONSCIOUS MENTAL FUNC
You are driving along in your car at 30 mph on a city

denly, a young boy steps into the street in front of you
ing a ball. You slam your foot on the brake pedal to b
to a screeching halt. Were you consciously aware o
before stepping on the brake? Or was that an unconsc
that you became aware of after you hit the brakes?
Unconscious Mental Functions

Our experimental evidence, described in Chapter 2, s
activations of the sensory cortex have to proceed for u
500 msec to produce awareness of a sensory sig
the duration of the liminal stimulus to the sensory
reduced below that threshold—such as to 400 ms
450 msec—no sensory awareness was reported. The
ported, “I felt nothing.” A similar situation was foun
of stimulus pulses applied to the specific ascendi
pathway in the brain; this is the fast pathway from t
to cerebral cortex.
In spite of this presumed actual delay in the aware
150 msec
Subjective antedating
Report
seeing boy
Fig. 3.1. Sequence of events when a boy steps in front of a movin
boy and the ball for up to 500 msec, you are capabl
ming on the brake in about 150 msec or less after th
peared (see Fig. 3.1). That action, therefore, must be
unconsciously, without awareness. Amazingly, you
awareness can be automatically but subjectively anted
ferred back in time, so you would report seeing the bo
ately (see Chapter 2, “Antedating of Delayed Senso
ence”).
Stepping on the brake is not a simple spinal reflex.
recognizing the nature of signal (in this case, a boy) a
sion to act, in order not to hit him. This fairly comp
function is carried out unconsciously.
Perhaps we should clarify what we mean by un
(nonconscious) functions and how they differ from
mental functions. The prime feature of a conscious ex
awareness. This is a subjective phenomenon, accessib
the individual with the experience. To study awarenes
rely on the person’s ability to indicate he has had such
U N C O N S C I O U S A N D C O N S C I O U S M E N TA L F U N
considered elsewhere in the book (see Chapter 1, “A
and the following section “Experimental Test of Tim
ory”). We regard a psychological function or event
scious when the person has no reportable awaren
event. This definition covers various possible kinds o
unconscious processes, from the depths of general an
so-called subconscious ones.
Dreaming is clearly a conscious process, even thou
tents may contain distorted events. Dreams are usu
recalled or not recalled at all. Dreams are thus ex
awareness with little or no memory.
Many of the routine functional activities in the
spinal cord) are carried out unconsciously. These incl
tion of blood pressure and heart rate in relation t
of exercise or emotional events; initiation and contro
ing; postural adjustments of the body and limbs; w
running; control of gastrointestinal motility and
regulation of endocrine glands (which secrete horm
even significant controls of the immune system. The
maintenance functions occur at rest and in locomo
proceed while we eat, during sexual activities, duri
flight responses, and so on. However, these kinds o
are not properly referred to as mental or psychologi
most of these activities do not and cannot ever reach
awareness.
Nevertheless, unconscious functions that do invo
logical or mental features are voluminous (see Velm
Conditioned reflexes can be learned without the subje
92 •
U N C O N S C I O U S A N D C O N S C I O U S M E N TA L F U N C T I O N S
subject learns to blink in response to the tone alone
nent psychologist Richard Thompson tells me he
tioned the eyelid blink reflex without any awareness of
tioning stimulus by the human subject. Other psycholo
also reported this phenomenon. Indeed, in a recent p
and Squire (1998) report that if the conditioning tone e
1,000 msec before the puff of air, the subjects can lear
ditioning only if they are aware of the stimuli. This 50
terval fits well with our evidence for the brain’s requ
produce awareness. Even the novelty of a change, in
learned sequence of stimuli, can be responded to w
awareness of that novelty by the human subject (Be
1997). Our quick reactions to sensory signals appear
formed without any initial awareness of the signal.
Another experimental example is provided in stud
jects reaching for a target that can shift its position. Ea
was asked to reach suddenly for a given target image
After the reaching movement began but before the
touched, the location of the target was shifted. Th
altered the direction of their reaching movements
stream” to touch the target at its new location. The
point here is that the subjects were not aware of m
midstream change in direction; the alteration was do
sciously ( Jeannerod, 1997).
Initiation of voluntary acts can arise unconsciou
brain, well before any awareness of any conscious in
move (see Chapter 4). That is, the brain starts the volu
cess unconsciously.
man, 1980). Much of this evidence deals with findings
son shows some mental effects of events even wh
no awareness of the events when they occurred. Fo
words or drawings were displayed on a screen so brie
10 msec) that the subject had no awareness of thos
drawings. But when asked later to respond to othe
drawings, the subject’s answers were clearly influen
earlier unconscious exposures.
There is also the commonly experienced phenomen
ing an intuitive feeling or hunch about an issue o
These hunches are based on unconscious, not consc
tal backgrounds or perceptions. Haven’t you had su
hunches? In a recent direct study of the hunch ph
Damasio et al. (1997) found that normal subjects co
quickly learn to pick cards that “paid off ” from th
cards that provided a successful payoff (two out of f
Subjects began to favor the money-making decks w
they could articulate any conscious strategy for th
Lowered skin resistance (presumably due to swea
creased blood flow in the skin), a bodily indicator of
peared during the time period in which the unconsc
was developing. This ability was found to be absen
tients with damage to certain areas of the frontal lo
cerebral cortex). General intelligence and memory w
paired in these patients; instead, they were specifical
in the ability to make successful decisions, and to do
sciously! Nichols and Newsome (1999) have reviewed
dence from experiments with monkeys. In these ex
94 •
is evidence that conversation or suggestive comment
duce effects on “subsequent thoughts, feelings and ac
out explicit recall of the events of surgery” (Bennett e
These unconscious processes have been confirmed by
of anesthesiologists. Of therapeutic importance is th
tion that the nature of the surgeon’s comments, receiv
sciously by the anesthetized patient, can influence the
later recovery. Positive statements can facilitate the l
ery while negative ones can disturb the recovery.
Many of our thought processes are apparently un
particularly related to attempts to solve a problem. T
cially striking in dealing with a problem in mathema
scribed by some of the great mathematicians. For ex
famous mathematician Henri Poincaré (1913) was in
how he arrived at mathematical solutions, and he
about the process. He wrote that he was puzzled abo
solve a particularly difficult mathematical problem
some conscious mulling over the problem, he gave up
a later trip to Lyon, the entire solution “popped” int
sciousness just as he stepped off the bus. Clearly, a gr
unconscious yet creative thinking had gone on to pro
solution. It is also said that when mathematicians c
perceive a solution to a difficult problem (this solution
after an unconscious process), they “know” intuitivel
that solution is correct, before consciously working o
essary analytical proofs. The case for unconscious me
tions was also made by the renowned mathematician a
opher, Alfred North Whitehead.
tive hypotheses of solutions to problems, which cons
peared only after some period of unconscious incubati
some of them describe an almost stereotyped seque
ducing new and original ideas of solutions: (1) specif
tions or problem; (2) gather or produce relevant i
about the issue; (3) suspend further conscious attem
duce an hypothesis that might lead to an answer (in o
let the concern about the issue percolate at an u
level); (4) be attuned to the conscious appearance of a
pothesis of solution; and finally (5) apply a conscious ra
ysis of what has finally arisen to a conscious level
usefulness and validity. Step (3) is probably the mo
one; the other steps are more in the nature of logic
Poincaré argued that intuitive work is even more imp
analytical work for the advancement of science. He
logic does not lead to anything but tautologies; it cr
ing new” (as quoted by Rafael Franco, 1989).
Creative ideas have also been reported to appear
and during daydreaming. Ideas and other thoughts t
in dream states are clearly not the result of a delib
scious analysis or process. They appear with no imm
thought and may be regarded as unconscious develop
pop into conscious awareness in the dream. There is
story of Otto Loewi, who was awarded a Nobel Priz
lishing that a chemical substance could be the transm
(synaptic) junction between one nerve fiber and the
or muscle cell. Loewi was frustrated by his inabilit
way to solve this question experimentally. Then on
96 •
ing night, Loewi arose, went to his laboratory, and
imagined experiment immediately. The originality o
lay in collecting the fluid coming from one frog’s hear
ing it into the intake of a second heart. When Loewi
the vagus nerve to the first heart, producing a slowin
porary stop of the first heart’s beating, the second
exhibited a slowing of its rate. The second heart c
have received the “message” from the first heart by
substance, released from its stimulated nerve and tran
the second heart. In subsequent years, such a chemica
ter agency was found to apply to most other junction
one nerve cell and the next, including those in the bra
nal cord.
Creative writing, painting, musical composition, a
mance are also widely believed to involve unconscio
processes, but I shall not attempt to elaborate this view
thur Koestler (1964) developed this view in The Art of
I have myself experienced numerous occasions in w
ductive new ideas have popped into my conscious mi
was dreaming or daydreaming. For many years, I
keeping a pad of paper and pencil at my bedside. Wh
up during the night with a novel idea, I make notes f
daytime action. A number of interesting solutions an
tions for research problems have appeared from tha
sometimes daydream when reading a book, taking a w
ing to instrumental music, or even while listening to
For example, while I am listening to a live sympho
mance, my mind often drifts off and other thoughts a
even in the darkness of the concert hall.
An example of a creative product of my daydream
idea for using a reportable clock method to establish w
son becomes aware of voluntarily wanting to move
ter 4). That occurred to me while I was sitting in my s
at the Rockefeller Center in Bellagio, Italy. I was supp
concentrating on writing a research paper on a quite
sue, one dealing with subjective antedating of a con
sory experience (Libet et al., 1979). The problem, of h
mentally to deal with the relation between brain pro
the conscious will to act, had reappeared the previo
ing a discussion with my wife Fay about the apparent
ity of solving that problem. The solution that popp
mind was to instruct each subject to associate the firs
of an intention to act with the position of the second
clock. The subject’s later report of that associated
could indicate the time at which the awareness appea
It is, in fact, important to allow your unconscious m
cesses to develop ideas and solutions, and to provide
ties for such processes to occur. Also, you must lear
nize and pay attention to a product of these proces
pops into conscious awareness. That is, you should
free flow of unconscious processes and learn to trust
portance. Such processes are often not stereotypic
they are often creative. When you become aware of
can make conscious choices about how to use or
them. Alfred North Whitehead urged everyone to c
98 •
Bower, 1999).
The importance of daydreaming for generating cre
and solutions is hard to get across convincingly to o
wife often felt I was wasting my time and not “work
she observed me sitting at my writing board and n
much. I finally convinced her (I think) that such app
tivity was not a complete waste.
Are Unconscious Functions “Mental”?

I have thus far avoided getting into a discussion o
“mind” and what is a “mental” process. You can find v
rate arguments on this topic in the literature, mostly b
phers. As an experimental neuroscientist, I tend to tak
direct approach that is in consonance with our report
and feelings about such concepts. According to diction
tions, “mind” refers to one’s intellectuality but also to
nations and impulses; in the latter sense, emotional
are included.
“Mental” is simply the adjective for describing fu
“mind.” Mind would thus subsume conscious experi
unconscious functions that fit into the definition wo
excluded. “Mind” might then be usefully regarded as
property of the brain that includes subjective conscio
ences and unconscious psychological functions.
But such a view has been strongly opposed by som
pher John Searle (1993, p. 156) argues that “mental” sh
unconscious mental event. He agrees, however, that t
ties could affect subsequent conscious thoughts, feelin
havior.
Well, then, why should we think of an unconscio
logically significant process as a “mental” process?
adopt that view, we are imparting an attribute to
scious process that makes it qualitatively akin in som
conscious process, except that it lacks awareness. B
(unconscious as mental or nonmental) are unproven h
But there are reasons for regarding the unconscious
feature, as one that better describes the known attrib
conscious functions. It also provides a more imaginat
tentially conjectural picture for dealing with these fu
Unconscious functions deal with psychological issu
that seem basically similar to conscious functions exc
absence of awareness. Unconscious functions can be
tions of experience (Kihlstrom, 1993). Cognitive, im
and decision-making processes all can proceed uncon
ten more creatively than in conscious functions. U
psychologically significant functions of these kinds
scious ones, cannot be described or predicted by
knowledge of the neuronal processes, contrary to Se
It seems simpler, more productive, and more in tune
cal experience to regard unconscious processes as “m
tions,” phenomena that are related to the conscio
functions but lack the added phenomenon of aware
all, definitions are only useful insofar as they promo
tive thinking about the item.) To add awareness to
100 •
Time-On Theory: How Does the Brain Distingu
between Conscious and Unconscious Mental Fu
Conscious and unconscious mental functions differ m
tantly in the presence of awareness for the former a
sence of awareness in the latter. We found that the
quires substantial time (about 0.5 sec) to “produce” aw
a sensory signal, while unconscious functions appear
much less time (100 msec or so). What was the brain
ing the shorter periods of activations that did not
enough to produce awareness? Far from being silent
exhibited recordable neuronal responses that resem
that went on to finally become adequate for awaren
shorter-lasting trains of nerve cell responses could no
awareness. But, we asked, could they provide a mech
an unconscious detection of a sensory signal? That qu
us to propose a time-on theory for explaining the tra
tween brain activities required for unconscious menta
and those required for conscious functions.
The time-on theory has two simple components:
(1) To produce a conscious sensory experience
words, with awareness), appropriate brain activities
ceed for a minimum duration of about 500 msec
event is near threshold). That is, the time-on or durat
activities is about 0.5 sec. We had already established t
experimentally.
(2) We proposed that when these same brain acti
durations shorter than those required for awareness,
the duration (time-on) of the appropriate brain activi
alized that time-on was probably not the only factor
sition between unconscious and conscious, but we
controlling factor.
You may ask, what is it that makes some tim
enough for awareness and most of the others not lon
We don’t have a full answer to that. However, there i
son to believe that focusing attention on a given sen
may be an agent for making the sensory response a
one. We don’t yet know what brain mechanism “dec
cus attention on one signal and not on others. But t
dence that the attention mechanism could “light up”
some areas of cerebral cortex; such an increase in
level of those areas might facilitate their lengthenin
tion of their nerve cell responses to achieve the t
awareness.
We don’t know precisely which neuronal activities
priate,” for either a conscious or unconscious mental
my argument is that whatever the appropriate nerve
ties are, the duration of those activities may be a critic
determining the distinction between the two kinds
events.
Experimental test of time-on theory. Any theory prop

entific must be testable. So we designed and carried
perimental test of the time-on theory (Libet et al.,
test involved two features: (1) We had to be able to v
ration of proper repetitive activations of the sensory
102 •
(2) We required a psychological task for the subject
whether the input signal was or was not “perceived”
regardless of awareness of the signal. That would a
match the duration of stimulus with both the accurac
tion of the signal and the level of awareness that it did
produce in each trial. Any correct detection of the s
no awareness of it would, of course, constitute an un
detection of the signal.
The first condition was met by applying stimulu
the ascending sensory pathway in the thalamus, belo
sory cortex. As shown earlier, with threshold limin
ties, minimum durations of up to about 500 msec wer
here also to elicit a conscious sensation. (We made
somewhat stronger than the minimum threshold, s
mately 400-msec durations were needed instead of
The actual duration of each train of 72 pulses per s
different for each testing trial, randomly ranging fr
stimulus) to about 750 msec (that is, from 0 pulses to 5
this setting). A train duration of 500 msec would
pulses here.
The subject faced a panel containing two button
which could be lit up briefly (see Fig. 3.2). In each tria
(L1) was lit for 1 sec; and 1 sec later, L2 was lit for 1 sec.
ulus to the sensory thalamus was applied in a rando
either during the time L1 was lit or during the L2 light
The subject’s task was to indicate in which of the tw
ods, L1 or L2, the stimulus was delivered. She had to
decision even if she were not aware of any sensation
#1
#2
#3
L1 L2
Fig. 3.2. Test of time-on theory. A box with this surface panel
subject. In each trial, light #1 (L1) lit up for 1 sec; then light #2
for 1 sec. Stimulus was delivered during either L1 or L2 randomly
sive trials.
After the L2 light was finished, the subject indicated whether
lus was delivered either during the L1 or the L2 light period by pr
L1 or L2 button. Subject was instructed to do this even if he o
nothing in L1 and L2.
Subject then also indicated the level of awareness of the st
pressing button #1, #2, or #3. Button #1: The subject felt the
even if weakly. Button #2: The subject had an “uncertain feel
stimulus; or, maybe something different,” even if not the same
felt for #1. Button #3: The subject felt nothing and was guessing
ing L1 vs. L2. From Libet et al., 1991. Reprinted with permission
ford University Press.
stimulus: Button #1 if she felt it, even weakly; #2 if sh
certain whether she felt it or if she felt anything differ
the selected light; #3 if she felt nothing and was just g
choosing L1 or L2.
The choice of L1 or L2 should, by pure chance, pro
rect responses in 50 percent of trials. Correct respons
than 50 percent of trials, with a given stimulus durati
indicate actual detection of the stimuli with that
whether with or without awareness of the signal. Ea
performed in hundreds of trials so the results could
cally analyzed.
The results were very instructive: (1) For trials in
stimulus (0 pulses) was delivered during either L1 or
sponses were indeed very close to 50 percent corr
pected from chance alone. (2) For all trials in which
was delivered, but subjects were not aware of any sen
were guessing, the correct responses were significan
than 50 percent. This was true even with short train du
15–150 msec (1 to 10 pulses). With longer stimulus tra
260 msec) and a guessing subject, subjects were 75 pe
rect, and so on. Clearly, subjects were often detecting
lus and making a correct response with no awareness of
of the stimulus.
(3) By a statistical analysis, we determined the dif
stimulus durations between condition A (correct resp
guessing and no awareness) and condition B (correct
with some minimal evidence of awareness, at the unce
In both of these groups, A and B, all responses we
B (correct, with minimal awareness) required an addi
ulus duration of almost 400 msec. In other words, t
awareness to a correct detection required an increase
duration of almost 400 msec for the repetitive train
This result was precisely as predicted by the time-on
The results proved that awareness is a phenomenon
of content. With content the same (correct report on p
stimulus), an increase of 400 msec in stimulus duratio
essary to add minimal awareness to the response. T
requirement for awareness per se makes it a functio
from others in the brain.
The results also provided direct evidence for a for
liminal perception.” Shorter durations of cortical
were subliminal in the sense that they did not prod
ness of the signal. Nevertheless, these subliminal inpu
sponded to correctly at well above the 50 percent cha
consider the potential implications of this finding for
perceptions generally in the following section (view
case, this result directly demonstrates the important
between unconscious detection of a signal and the cons
ness of a signal.
The transition between unconscious detection and
awareness of the stimulus was here produced simpl
able increase in duration of identical cortical activatio
direct ascending sensory pathway). The result gave us
fidence in the theory, and permits us to speculate on
portant implications that follow from the time-on the
106 •
Recall that in the time-on theory, the feature that adds
to an otherwise unconscious psychological function
stantial increase in the duration (time-on) of the a
neuronal activities. The theory thus suggests or leads
lowing views.
(1) Perhaps all conscious mental events actually be
sciously before any awareness appears. We already ha
perimental evidence that this situation occurs in th
awareness of a bodily sensation, and also for the inte
erated awareness of the intention to perform a vol
(see Chapter 4). That is, to elicit any such awareness
substantial duration of cerebral activities. That mean
conscious, shorter-lasting cerebral activities have pre
delayed conscious event. It seems likely that such a fu
requirement, found by us for two different kinds of
experience, would also apply to other kinds of awa
other words, for the other sensory modalities (vision
smell, taste) and for conscious thoughts and feelings,
or otherwise.
Application of such a principle to internally genera
ing and emotional feelings introduces a very intere
bute. Thoughts of various kinds, imaginations, atti
ative ideas, solving of problems, and so on initial
unconsciously. Such unconscious thoughts only reach
conscious awareness if the appropriate brain activities
enough time.
cal change (the readiness potential or RP) that beg
sciously in a simple voluntary action also precedes
voluntary actions of speaking or writing (see R. Ju
discuss the impact of this finding for the nature of con
in Chapter 4. In the case of speech, for example, this
the process to start speaking, and even the content o
be spoken, has been initiated and prepared unconscio
the speaking begins. If the time-on requirement for
holds here, it would be manifestly impossible to rapi
series of words, in the usual fashion, if one first had
consciously aware of each word. When a spoken wo
thing different from what the speaker would conscio
have said, he usually corrects that after hearing himsel
deed, if you try to be aware of each word before spea
flow of your speech becomes slow and hesitant.
In smoothly flowing speech, words are allowed to
their own,” in other words, they are initiated uncon
E. M. Forster reportedly stated, “How can I tell what
til I see what I say?” Then there is the event rec
Bertrand Russell after a late night talk with Lady Ot
sell wrote, “I did not know I loved you till I heard my
you so—for one instant I thought, ‘Good God, what h
and then I knew it was the truth.” (These two exam
described in a paper by Sean Spence, 1996.) And, t
elegant statement by writer E. L. Doctorow, “I lo
my mind flowing through sentences and making dis
trust the gift of writing and see what it will deliver
My daughter Gayla tells me that when she writes
108 •
(3) The playing of a musical instrument, like the pia
lin, or singing must also involve a similar unconscio
mance of the actions. Pianists often play rapid music
which the fingers of both hands are hitting the keys in
so fast that they can barely be followed visually. Not
each finger must hit the correct piano key in each se
would be impossible for a pianist to become conscio
of each finger’s action if there were a substantial de
awareness of each finger’s movement. Indeed, perfo
port that they are not aware of the intention to act
finger. Instead, they tend to focus their attention on
their musical feelings. Even these feelings arise uncons
fore any awareness of them develops, based on ou
principle for producing awareness. Instrumentalists a
know that if they “think” about the music being p
their expression becomes forced and stilted. Smoothly
music, with heartfelt and spiritual feelings, is produced
performer allows the expression to arise without cons
ifications, in other words, to arise unconsciously. Musi
close their eyes while performing; perhaps that action
getting in touch with their unconscious feelings whil
external signals. I have first-hand reports of all the
from my four children, who are advanced string pl
from my own singing experience.
(4) All quick behavioral, motor responses to a sens
are performed unconsciously. These are responses th
made within 100–200 msec after the signal, well bef
ness of the signal could be expected. Many actions in
1 2
time
Fig. 3.3. Timings for baseball batter’s response to a pitched ba

pitcher delivers the ball at 90 mph with curved trajectory. The b
the batter in 450 msec. The batter may wait until the final 200 m
ing the #1 period) to try to detect the course of the ball’s mov
him.
The batter must decide whether to swing (#2 period) before t
msec or so; the 150 msec is the minimum time required to activa
tor cortex, which sends a neural message down to the spinal mo
cells that activates the appropriate muscles in about 50 msec;
muscle contractions that produce the swing of the bat occur in
msec. The great homerun hitters, like the current champion Bar
can swing the bat with extraordinary speed. That enables them
their decision to swing until the minimum required time to hit t
into this category. A professional tennis player must re

ball served to him at 100 mph and with a curving
These players report being aware of the serving mot
of the opponent, but they are not immediately aw
ball’s location when hitting the return. A baseball ba
pitched ball at 90 mph (132 ft/sec), with the ball curv
ing at the last moment. He must decide whether to h
and to swing the bat in a path that can meet the ba
3.3). Because the pitcher is 60 ft from the batter, the b
110 •
recognition and decision are presumably initially un
Great baseball hitters are probably those who can su
delay these processes as much as is physiologically
Once the baseball batter has made his decision and
swing, it is remarkable that he usually cannot stop h
he realizes it was the wrong choice.
I might even add that great athletes, in general, are
can let their unconscious mind take over without in
from the conscious mind. Athletes tell us that if th
think” (become aware) of immediate responses, the
less successful. Indeed, I am tempted to generalize t
true for all creative processes, in art, science, and mat
Quick responses to signals can be measured quant
reaction times (RT) studies. In an RT study, the actual
are presumably made unconsciously, with awareness
nal following the act. It has, in fact, been shown tha
time to a given signal can be the same even when aw
the signal is completely obliterated. This obliteration c
duced by applying a delayed masking stimulus foll
initial signal for which reaction time is measured (T
McCloskey, 1990).
(5) Unconscious mental functions can proceed
speed, if they are carried out by shorter-lasting neuro
ties. Judging from our experiments on signal dete
forced choice responses with no awareness of the sig
fective time-on for neural activities in unconscious fun
be very short indeed—about 100 msec or less. This im
the series of unconscious processes involved in solvi
lem can proceed speedily, each brief process after ano
plex problem. By contrast, if a person did not pro
awareness appeared for each step in a series of tho
whole process would be slowed down by a factor of
and conscious thoughts and resulting decisions to ac
come a plodding affair.
(6) The appearance of a conscious experience ha
nothing character (see Fig. 2.2). That is, there is no
conscious awareness of an event even if the appro
ronal activities persist for as much as 90 percent of th
required for actual threshold awareness. What the t
periments have demonstrated is that threshold awar
in rather suddenly when the activities persist for th
msec requirement!
(7) The popular notion that people have a continu
of consciousness is contradicted by the time-on requ
conscious awareness. The notion of a stream of con
was proposed by the great psychologist William Jam
basis of his intuitive grasp of his own conscious thou
psychologists and writers of fiction have adopted the
stream of consciousness as an authentic characterist
ject’s or character’s mental activities. But our evidenc
that conscious thought processes must consist of disconti
rate events. If the beginning of each conscious event o
after a substantial delay, incurred by a required 500-m
of neuronal activations, then a series of conscious ev
not appear in a continuous stream. Awareness in each
event is not present in the initial 500 msec or so.
Discontinuity in a series of conscious events is
112 •
automatic subjective referral of each experience back
the fast-evoked response of the sensory cortex, a res
occurs within 10–20 msec of the sensory stimulus. Su
we do not perceive any appreciable delay in our aw
sensory events. Our experiment showed that peopl
they were aware of a sensory stimulus about 500 m
they could possibly have become aware of the stimulu
crepancy became known to us objectively; it is no lon
retical speculation. We called this phenomenon “sub
ferral of conscious sensory awareness backward in
Chapter 2).
However, this feature cannot be applied to all othe
conscious experiences, including the conscious inten
and thought events generally. We (Libet et al., 1979)
subjective antedating (referral backward in time) on
sory experiences. Even in that case, antedating occurs
the sensory input elicits a fast timing signal in the se
tex, the primary evoked potential (see Chapter 2). In t
the endogenous appearance of a conscious voluntary
to act, we have experimentally demonstrated that the
timing of that experience is in fact delayed by about 40
more after the onset of the brain activity that leads to
tary act (Chapter 4). Conscious intention to act, with n
cues to instigate it, is an example of a conscious expe
arises within the brain (in other words, endogenously
no primary evoked potential here, as there is in the se
tem’s responses to stimuli that are not endogenous in
Perhaps our subjective feeling of a smooth flow in
0 0.5 1 2 3 4
Fig. 3.4. Overlap of discontinuous mental events and the fee

smooth stream of consciousness.
M-1 conscious mental event begins suddenly after a 500-msec
unconscious initiating processes. M-2 conscious mental event ma
ter its unconscious initiating processes but before the end of M-1
for M-3 and M-4.
The overlapping of the successive conscious mental even
breaks in the stream of consciousness.
thoughts is explainable by an overlapping of the differ

events (see Fig. 3.4). The brain appears capable of s
scious events occurring almost simultaneously, ove
time. To illustrate how discontinuities of underlying
still produce a smoothly continuous overall product, t
the physiology of muscle action. A skeletal muscle,
ceps in the arm, is made up of many motor units, ea
ing many individual muscle cells or fibers. When you
smooth contraction of the biceps, like bending or fle
bow, an electrical recording of any single motor un
would show that it is “popping off ” at a relatively l
roughly ten times a second. Direct studies of indivi
responses show that the muscle contractions at ten
second are jumpy or wave-like, not smoothly sustain
tions. The smoothly sustained overall contraction of
biceps muscle is therefore explained as the result o
114 •
phase in one unit is accompanied by a contraction ph
other, and so on. If we electrically stimulate the wh
nerve to the biceps at a rate of ten per second, we ca
the motor units to respond synchronously at this rate
chronized contractions at ten per second in fact do
jumpy, tremor type of contraction of the whole bicep
(8) The time-on requirement for conscious experi
serve a “filter function” to limit conscious experien
one time. It is clear that very few out of the thousan
sory inputs delivered per second to the brain achieve
awareness, though they may lead unconsciously to m
cerebral and psychological responses. The French p
Henri Bergson proposed that the brain may block mo
inputs from access to consciousness to protect us f
overwhelmed by conscious responses to them. Our p
perimental findings may provide a physiological mec
achieve that blocking.
We propose, then, that the large majority of sens
remain unconscious because they do not develop a s
long duration (time-on) of the appropriate cerebral
activities. Perhaps it is the attention mechanism tha
given selected response to last long enough to elicit
but attention itself is apparently not a sufficient mech
awareness. Thus, the time-on requirement for aware
provide part of the mechanism for screening out senso
which do not reach awareness.
The screening or filtering of inputs prevents consci
ness from becoming cluttered and permits it to be f
ders reflect improper functioning of such a filter mec
an abnormal reduction in the duration of brain a
quired for awareness.
(9) Unconscious detection of a signal should be cle
guished from conscious awareness of the signal. This
was directly demonstrated by the results of the expe
viously described as a test of the time-on theory. But
tion is often overlooked, leading to confusing and
conclusions about the nature of conscious experien
based on signal detection theory find that subjects re
rectly, in other words, at better than chance levels,
signals with intensities down near zero. This has led
clusion that there is virtually no threshold level r
elicit a (conscious) sensory perception; accuracy of re
creases smoothly along a curve relating accuracy to s
tensity, the latter starting from zero. That conclusio
matches our results for unconscious detection of sens
without awareness. In studies of signal detection (
Swets, 1966) and of many other psychophysical issu
are asked to make a forced-choice response. In a for
the subject is asked to answer “yes” or “no” to a que
the stimulus; she is not asked about whether she wa
the stimulus. The two different questions can produc
different results.
The forced-choice question, strictly speaking, stud
of a signal, whether unconscious or accompanied by
A couple of interesting examples illustrate this: V
(1984) found that, for sensory input from the skin, th
116 •
or “no” as to whether some sensory message was
Vallbö himself has agreed that this did not reflect sens
ness, and that it was probably a case of unconscious sen
tion (personal communication). But many neuroscie
been erroneously regarding his finding as an indicato
solute potentiality for conscious sensory perception.
Human subjects can discriminate between two
stimuli of different frequencies to the skin. This can o
when the time intervals, between the individual repet
tory pulses, are much less than the 500-msec time
have found necessary for threshold awareness of
event. It has therefore been argued, by some, that ou
for the long time to awareness cannot be correct, b
can tell the difference between vibrations that each h
shorter intervals between pulses. But the ability to di
between short time intervals between vibratory pulse
ent frequencies demonstrates a detection of these d
awareness of that discrimination comes later, in our
is, my question is, when is the subject aware of the d
tion, not how short an interval between pulses can he
The reports of blindsight by Lawrence Weiskrantz
vide a beautiful example of the distinction between un
detection and conscious awareness. Patients were stu
had lost their conscious vision in some part of their v
due to damage of the visual cortex. When asked to
target in that blind area, even if they were guessing, t
with remarkable accuracy but reported they could n
target.
nal stimulus. Direct evidence for this appeared in
mental test of the time-on theory (see discussion ea
chapter). Subliminal perception is less easily proven
nary natural sensory stimuli are used. This is because
ences (of strength, duration, and so on) between a
and a supraliminal (awareness-producing) sensory st
usually small. However, a considerable amount of i
dence supports the existence of subliminal percepti
mostly deal with the alterations in later tests, applied
posure to stimuli whose content did not reach consc
ness. The subject’s responses to the later tests show a
of the previous subliminal stimuli that themselves pr
awareness. In an early study, Howard Shevrin (1973)
sual drawings or words so briefly (1–2 msec) that t
were completely unaware of the content in that flas
tests showed that these subliminal contents had an e
subject’s choices of responses in tests of word assoc
subjects remained unaware of these effects. Many o
gous tests have been reported in which subliminal w
“primed” the later responses of the subjects in test si
(11) Where in the brain do unconscious and cons
tions take place? Are there different locations for th
pects of mental function? The time-on theory sugge
conscious and conscious functions could both be m
the same brain areas, by the same group of neurons. If
tion between the two functions is simply one of a lo
tion of similar nerve cell activities to elicit awarenes
not postulate separate neuron entities for each. It is,
118 •
case, the single area with time-on control may not rep
only distinction between unconscious and conscious
However, the time-on feature could still be a control
for the distinction, in whatever areas of the brain it is
The phenomenon of blindsight(see Weiscrantz, 19
the possibility of separate pathways and brain structur
scious and unconscious functions. A human patient w
in the primary arousal area of the cerebral cortex is bli
he has no conscious vision for the external visual fi
normally represented in the area that is destroyed. Ne
such patients can correctly point to an object in that
when asked to do that simply as a forced choice. The s
port that they do not consciously see that object.
The unconscious blindsight action may be carried
area or network in the brain that is different from th
scious vision (for which the primary visual area is n
However, in an alternative explanation, both the con
unconscious visual functions may “reside” in some
outside the primary visual cortex, for example, in som
ary visual area. The function of the primary visual co
then be to fire inputs repetitively to this secondary are
increasing the duration of the activities there so as to
ness to the visual response. This effect would be ab
the primary visual area is not functional.
Can you have conscious perception without the p
sual area (V1)? In a very interesting study, Barbur e
claimed to have shown that one can. They studied a p
had lost area V1 completely, by damage in a car accide
through his verbal reports, that he is consciously aw
the nature of the visual stimulus and its direction of
However, Barbur et al.’s conclusion—that consc
perception is possible in the absence of V1—does n
our time-on theory. It is possible that area V5, which
creased activity in response to the visual stimuli, m
production of visual awareness by virtue of a suffic
duration of activations. Indeed, Barbur et al. (1993) d
sual stimuli repetitively during substantial periods of
(12) Modulation of the content of a conscious ex
recognized as an important process in psychology a
try. It is most directly demonstrable when a person
experience that differs from the actual visual image
Persons who are emotionally disturbed by the sight
woman may report seeing an altered version of the n
shown to them. (An eminent Swedish neurologist w
he had tried this particular example on his subjects.
that nude pictures would not be adequate as psycholo
blemakers in Sweden.) The alteration in content of
ence appears not to be one of conscious distortion; th
unaware of his distortion of the image and the proc
to be an unconscious one.
Freud, of course, made use of the modulatory ph
in his views of the unconscious effects of emotion
on a person’s conscious experience and verbal expre
Shevrin, 1973). The time-on theory provides a physio
portunity in which unconscious modulations of the
an experience can occur. To effect a change in the
120 •
ation of the conscious image. During the time inter
conscious sensory awareness appears, brain patterns
tect the image and react to it, by producing activities th
the content of the conscious experience before it appe
Our evidence indicates that a substantial period of
tivity (500 msec of time-on) is in fact required to elicit
of the sensory event. That delay provides a simple and
physiological opportunity during which unconscious
terns can alter the content of the experience before
of it appears! Indeed, the experimental phenomenon
tive referral of a conscious sensory experience backwa
provides relatively direct evidence for one kind of m
distortion of the subjective experience. The delayed
is subjectively timed as if it were not delayed at all. O
experimental findings showed that the subjective expe
skin stimulus could be reported to be distinctly stron
actually was, when that skin stimulus was followed b
cortical stimulus that started as much as 500 msec later
ter 2). That is direct evidence that the time period
in which the sensory experience is being finally bro
awareness can be used to alter the content of the expe
fore it reaches awareness.
Any modulations or modifications of the develop
ence would be unique to the person involved. It wo
the person’s own history of experiences and his emo
moral make up. But the modulations are made unco
Consequently, one may say that the unique nature
person can express itself in unconscious processes. Th
neuronal time requirement for producing awareness
on factor) can have penetrating impacts on our view o
in which a variety of unconscious and conscious m
tions operate. It is important to note that these neu
factors could only have been discovered by direct e
on how the brain deals with conscious experience, no
lative theories based on previous knowledge of brain
122 •
4
INTENTION TO ACT: DO WE HAVE FREE WILL?
How the brain deals with voluntary acts is an issue

mental importance to the role of conscious will an
that, to the question of free will. It has been commonl
that in a voluntary act, the conscious will to act wo
before or at the start of the brain activities that lead to
that were true, the voluntary act would be initiated an
by the conscious mind. But, what if that were not the
possible that the specific brain activities leading to a
act begin before the conscious will to act, in other wo
the person is aware that he intends to act? This pos
arisen partly from our evidence that sensory aware
layed by a substantial time period of brain activities. If
nally generated awareness of the will or intention to
delayed by a required period of activities lasting up to
msec, it seems possible that the brain’s activities tha
willed act begin well before the conscious will to ac
adequately developed.
We were able to examine this issue experimentally
found, in short, was that the brain exhibited an init
cess, beginning 550 msec before the freely voluntar
becomes aware of her will or intention to perform t
experimental evidence for this surprising sequence
this chapter.
Experimental Design
The possibility of an experimental investigation of th
was opened by a discovery made by Kornhuber a
(1965). They found that a recordable electrical chan
activity regularly and specifically preceded a volun
voluntary act was preceded by a slow rise in electrica
recordable at an area of the scalp located predomina
vertex, the top of the head. The electrical change sta
800 msec or more before a subject performed an app
untary act. It was therefore called the readiness potent
in German, the Bereitschaftspotential.
The act under study was a sudden bending or fle
wrist or fingers. Each RP is very small and virtu
among the other electrical activities of the resting br
fore, many such acts had to be performed to produ
puter-averaged tracing that summated the small RP
ject was allowed to perform these numerous acts
paced” manner. But his own choosing of times to a
ited by the period of about 6 sec that was allowed by
and Deecke for each trial, in order to achieve the sum
200–300 RPs within an acceptable experimental time
Kornhuber and Deecke did not consider the questio
the conscious will to act appeared, in relation to the b
124 •
I N T E N T I O N T O A C T: D O W E H A V E F R E E W I L L ?
of appearance of the conscious intention to perform
tary act. In a public discussion of willed actions, t
scientist and Nobel laureate Sir John Eccles stated his
an RP starting >800 msec before a voluntary act m
that the associated conscious intention appears even b
early beginning of the RP. I realized that there was n
to support Eccles’s view, which was presumably c
his own philosophy of mind-brain interaction (see P
Eccles, 1977).
Establishing the time of the conscious will relative
set of brain activity (the RP) was clearly important. If
will were to follow the onset of RP, that would hav
mental impact on how we view free will. But, at the t
no way to test the issue experimentally. It seemed im
achieve a valid measurement of the time of appearan
scious intention. Conscious will is a subjective phenom
directly accessible to external observations. It require
by the human subject who is experiencing that subjec
Having the subject press a button or say “now” to in
conscious intention would add further voluntary a
wrist flexion being studied. That would obscure the va
of the conscious will for the test act, relative to the bra
Also, there was no assurance that pressing a button
“now” as quickly as possible would be performed c
That is, the subject might make this quick respon
sciously, before becoming aware of the experience.
would not have a valid time for the conscious will.
While I was a Resident Scholar at the Rockefeller
I N T E N T I O N T O A C T: D O W E H A V E F R E E
time” for her experience of the conscious intention
clock time would be noted silently and reported afte
was over. Upon returning to San Francisco, we dev
technique (Libet et al., 1983).
A cathode ray oscilloscope was arranged to have
light revolve near the outer edge of its face. The ou
the oscilloscope tube face was marked in clock secon
usual, around the circle. The movement of the light s
signed to simulate the sweep of the second hand
clock. But our light spot completed the circle in 2.56
twenty-five times faster than the normal 60 sec (se
Each marked clock second therefore corresponded t
msec of the spot’s motion. This faster “clock” could
time differences in hundreds of milliseconds.
The subject was seated about 2.3 m from the oscillo
each trial, the subject fixed his gaze on the center of
scope’s face. He was asked to perform a freely volun
simple but sudden flexion of the wrist at any time he
ing so. He was asked not to preplan when to act
should let the act appear “on its own.” That would
separate the process for planning an act from that
spontaneous will to “act now.” He was also asked t
his first awareness of his intention or wish to mov
“clock position” of the revolving light spot. That
clock time was reported by the subject after comple
trial. We labeled these reported times “W,” for conscio
ing or wishing or willing to act. The RP produced in
voluntary act was also recorded, with suitable electro
126 •
Fig. 4.1. The “clock” for timing a mental event. A spot of light, g
in a cathode ray oscilloscope, is arranged to move around the ed
face of the oscilloscope, completing the circle in 2.56 sec. That
the sweep second hand of the usual clock, but moves about tw
times faster.
Numbers around the periphery represent seconds for the usu
sweep, but each marked second actually corresponds to about
here. Described in Libet et al., 1983.
head. An RP averaged over forty trials was found sui

time of onset of this averaged RP could then be co
the reported W times averaged for the same forty act
We had serious doubts initially that the subjects co
their clock times of conscious intention with sufficien
standard error (S.E.) of close to 20 msec. That was tru
subject, even though the averaged Ws differed amon
Because averaged Ws for all subjects was about −20
fore the motor act), an S.E. of ±20 msec provided ad
ability.
A test for accuracy of W was a bit trickier to devise
not know, in an absolute way, how close the reported
the actual subjective time of that awareness. But we
how accurately the subjects were using our clock
nique. For this, a series of forty trials was run in wh
skin stimulus was delivered to the hand. The sub
asked not to perform any voluntary act but rather t
clock time of the skin sensation, to be reported afte
(as for W). The skin stimulus was delivered at ran
times for the forty trials. These times (“S”) were, of
known to the subject, but they did become known to
ers in the computer printouts. We could thus compar
tively known expected time of a subjective awarene
clock times reported by the subject. The reported S
close to the actual stimulus times. But they did show
of about −50 msec (in other words, earlier) from the
livered stimulus times. Because this difference was fa
tent, it could be subtracted as a bias element from t
W of −200 msec. That produced a “corrected” avera
150 msec. A series testing reported times of a skin st
run in each session.
Our definition of a voluntary act included the follo
will to act arose endogenously. That is, there were n
128 •
to act as well as whether or not to act. Human su
distinguish this last criterion from situations in whic
act is produced without such a qualification. The neu
Wilder Penfield (see Penfield and Rasmussen, 1950)
stimulated the exposed motor cortex during surgery t
leptic foci. Stimulating the motor cortex produces co
of some muscles and some movements in specific s
body. The patients reported that they did not will s
ments; they reported that these actions were impose
by the stimulator, they were not voluntary.
There are numerous clinical disorders in which act
in the absence of conscious will. These include the in
actions in cerebral palsy, Parkinsonism, Huntington
Tourette’s syndrome, and even obsessive compulsion
striking example is the “alien hand sign” (Goldberg a
1990). Patients with a lesion in a fronto-medial porti
motor area of cerebral cortex may find that the han
on the affected side perform curious purposeful actio
undoing a buttoned shirt when the subject is trying to
up. All of this occurs without, or even against, the su
tention and will.
Two Groups of RP Onset Times

Our experimental objective was to study freely volu
performed with no external restrictions as to when
most of our series, each of forty trials, there were no
preplanning by the subjects. These voluntary acts w
lowed us to place recording electrodes on the actua
be activated; the recorded electromyogram gave us t
the act and also served as a trigger to the computer to
scalp potential that had appeared during the 2 to 3 s
the muscle activation. But the time of the act was
free for the subjects’ own will. Our experimental qu
Does the conscious will to act precede or follow the
tion? Testing this required only that the timing of the
freely up to the subject. The nature of the act was no
for that question.
For some series of trials the subjects reported
planned a range of clock time in which they would a
of our encouragement not to do that. Those serie
RPs (#I) with earlier onsets, averaging about −800
msec (before the motor act) (Fig. 4.2). These values w
to those reported by Kornhuber and Deecke and by
their “self-paced” movements. For this and other rea
peared that “self-paced” acts, done with certain limi
posed by the experimenter, probably involved some p
by the subject of when to act. Their subjects knew t
perform the act within 6 sec, and that may have e
some preplanning of when to act. Our subjects ha
restriction.
In those series of forty acts in which the subjec
no preplanning of when to act, the onset of RPs (#I
−550 msec (before activation of the muscle). It shou
that the actual initiating process in the brain prob
before our recorded readiness potential, RP, in an
130 •
Fig 4.2. Readiness potentials (RP) preceding self-initiated volun

Each horizontal row gives the computer-averaged electrical pot
corded with an active electrode on the scalp, either at the midlin
for subjects G. L. and S. B., or on the left side (contralateral to
forming right hand) over the motor/premotor area that controls
for subject S. S.
RPs labeled type II were found when every act (quick flexio
wrist) in the series of forty trials was subjectively reported to
spontaneously, with no preplanning of when to act. Type I RPs w
recorded when the subjects experienced a preplanning to act so
within the next second or so.
In column S, a near-threshold skin stimulus was applied in each
trials, at randomized times unknown to the subject. The subject w
to recall and report, after each trial, the clock time when he becam
of the skin stimulus. This was similar to the reporting for awarene
time of wanting to move. The large positive ERP (event-related-p
with a peak at about 300 msec after the stimulus, is commonly
when there is uncertainty about the stimulus (timing, in this case
The solid vertical line in each tracing represents “O” time, at w
activation of the muscle has begun (indicated by an electrom
EMG) in RP series, or the time at which the skin stimulus was de
the S series. From Libet et al., 1982. Reprinted with permiss
Elsevier, Electroencephalography and Clinical Neurophysiology, vol. 54
335.
recorded RP.
W values, for times of first awareness of the wish t
aged −200 msec for all series. (This could be correct
msec by the reporting error of −50 msec found for
stimulus) series.) W times were the same whether th
sociated with RPI or RPII. That is, W times were
whether there was or was not any preplanning of w
That indicated that the final volitional process (to
starts at about −550 msec; it is the same, whether fu
neous or preceded by deliberations or preplanning o
act. This final process may be the “act now” feature
tary process, and the events in the “act now” feature
regardless of preplanning.
The “act now” process should be distinguished fro
tions and advance making of choices about perform
One can, after all, deliberate all day and never act. W
study the deliberation phase of volition, except fo
sional preplanning by our subjects of when to act.
There have been questions about the meaning
times. Because we produced evidence for a delay
msec) for the development of a conscious sensory
awareness of the clock time may have started well
conscious W report. But our subjects were asked t
clock time associated with their first awareness of t
act; they were not asked to report the time when th
conscious of that association. There was presumably
up to 500 msec before the conscious time appeared;
tomatic backward referral or antedating to the init
132 •
our tests of reported times for a skin stimulus.
Robert Doty (personal communication) raised a di
tential source of error in the interpretation of our
This involved the “cost” in extra time for switching a
another task. The extra time required for switching ta
up to 100 msec or much longer in some instances. As
our case, “one cannot simultaneously attend to the int
world of decision (to act) and to the position of the
time) on the CRO (cathode ray oscilloscope).” Doty
gested that the subject’s free will starts the RP; when a
turned to the oscilloscope clock, there is the cost in sw
this task. That could result in a belated report of W fo
that actually was observed at the start of the RP.
My reply to the switching task argument is as follo
delay from the onset of RPII was, as corrected, 400 ms
longer than the usual cost of switching tasks, even if s
exists here. (2) The conditions in our trials were quit
from those in the reports of costs for switching tasks.
ter reports, the switch in tasks took place in complete
trials. In our case, the subject was already fully instruc
the trial. Our task, given in advance, was to watch fo
est experience of the urge or wish to act (W) while
the “clock” continuously so as to associate the occurr
with the clock position. This all happened in the same
task requirements that were different from those re
producing a cost for switching. (3) RPI (when there is
planning of when to act) had an onset of about −800
msec; RPII (with spontaneous unplanned acts) had a
Both kinds of trials involve similar tasks, and simil
switchings in tasks if that applies here. But then, you
count for the difference in RP-W intervals, in the m
gested by Doty. That is, you can’t have W actually s
RP, and appearing much later after onset of RPI than
if any costs in switching are the same for both cases.
the trials with skin stimuli (instead of movements)
eliminate the switching cost suggestion. The tasks in
lus series were essentially the same as those in volun
ries. Subjects were asked to monitor the “clock” spo
sociate the position of the clock spot when they fel
skin sensation produced by the stimulus at random
times in different trials. The subjects in fact reported
that were very close to the actual times of the stimul
the reports averaged −50 msec, relative to actual stim
This degree of accuracy does not allow for costs of s
the hundreds of milliseconds.
Sequence of Events in the “Act Now” Situation

So, what answer did we obtain to our original ques
relative timings for onset of brain activity (RP) versu
will to act? The clear answer was: The brain initiates
tary process first. The subject later becomes conscio
of the urge or wish (W) to act, some 350 to 400 ms
onset of the recorded RP produced by the brain. Th
for every series of forty trials with every one of the ni
134 •
precisely. Haggard and Eimer added interesting exp
features: They not only recorded the RP at the ver
did), but also the RPs exhibited by the lateral premoto
cortex. These lateral RPs (LRP) had onsets closer to
msec seen in our RPII recordings. Haggard and Eim
vided the LRP trials into a group with the earlier onse
with the later onsets. The reported W values (time of
of the urge to act) for the earlier LRPs were also the
values, and the Ws for later LRPs were in a later gro
ings. However, in both groups of trials, the onsets of
ceded the W times in the respective group. That sh
finding of LRP onsets preceding the W timings by
values is valid throughout the range of values for both
and Ws.
Haggard and Eimer made an additional point: The
that the vertex RP process cannot have a causal relat
appearance of W, because their RPs did not covary
early versus late Ws. But our RPII is the significant v
related to final initiation of the voluntary “act now
(RPI starts with a deliberation of when to move; tha
rate process.) Therefore, it is our RPIIs that should
into early and late groups, to test for covariance with
late Ws. This measurement has not been made, eithe
by Haggard and Eimer, and so no conclusion about ca
these grounds can be presently drawn (see Haggard
2001).
Philosopher John R. Searle (2000a and b) proposed
Therefore, a conscious self could not initiate that pr
reason for action developed by a conscious self wou
belong in the preplanning or choice-making categor
onstrated experimentally that that kind of process
different from the final “act now” process. One ca
plan and deliberate about an action without ever acti
philosophically generated models suffer from a failure
experimentally known evidence into account. His
mostly untested and even untestable.
Back to our experiment: An additional important fi
that W preceded the actual movement of the muscle
by about 150–200 msec (see Fig. 4.3). Also, the actua
between actual cerebral initiation and conscious w
probably greater than the 400 msec observed here
RP). As noted above, an unknown area elsewhere i
may be initiating the activity we record as RPII.
What does this mean? First, the process leading to
act is initiated by the brain unconsciously, well befo
scious will to act appears. That implies that free will,
would not initiate a voluntary act.
There are also broad implications for the timing
tary acts where speedy initiation is required, as in m
activities. A tennis player returning a ball served at 10
hour cannot wait to become aware of his decision
sponses to sensory signals in sports require complex
erations to meet each unique event. They are not ord
tion times. Even so, professional sports players will t
you are “dead” if you consciously think about your m
136 •
Fig. 4.3. Diagram of sequence of events, cerebral (RPs) and s

(W), that precede a self-initiated voluntary act.
Relative to “O” time (muscle activation), cerebral RPs begin fir
with preplanned acts (RP I) or with no preplannings (RP II). Subj
perience of earliest awareness of the wish to move (W) appears
−200 msec; this is well before the act (“O” time) but is some 350 m
even RP II. Subjective timings of the skin stimulus (S) averaged a
msec, before the actual stimulus delivery time. From Libet,
printed with permission from Cambridge University Press.
Conscious Veto
The finding that the volitional process is initiated unc
leads to the question: Is there then any role for con
in the performance of a voluntary act (Libet, 1985)?
scious will (W) does appear 150 msec before the moto
though it follows the onset of the cerebral action (
least 400 msec. That allows it, potentially, to affect
the final outcome of the volitional process. An inter
msec would allow enough time in which the consciou
might affect the final outcome of the volitional pro
tually, only 100 msec is available for any such effect. T
completion with no possibility of its being stopped by
the cerebral cortex.)
The conscious will could decide to allow the vol
cess to go to completion, resulting in the motor act it
conscious will could block or “veto” the process, so t
tor act occurs.
Vetoing of an urge to act is a common experienc
us. It occurs especially when the projected act is r
socially unacceptable, or not in accord with one’s ove
ality or values. In fact, we showed experimentally th
of a planned act was possible even during the last 10
before the expected time of the action. This was a l
It could not be done for a spontaneous veto, as th
no muscle activation electrically to trigger the comp
cord the preceding seconds of the electrical activ
scalp. So, we were technically limited to study a
act that was planned to be executed at a preset time
ject was asked to prepare to act at a certain time of t
say at the 10-sec mark. However, the subject was t
expected act when the clock reached 100–200 msec
preset time. A substantial RP developed during the
seconds before the veto, in accord with the subje
of feeling an expectation to act. But this RP flattene
100 to 200 msec before the preset time, as the sub
the act and no muscle response appeared. The ob
plied a trigger signal to the computer at the pres
act. This at least demonstrated that a person could
138 •
conscious will and of free will, in a volitional process
an act. Extrapolating our result to other voluntary
scious free will does not initiate our freely voluntar
stead, it can control the outcome or actual performa
act. It could permit the action to proceed, or it can
that no action occurs. Permitting the volitional proc
ceed, toward producing the motor act, could also inv
tive role for conscious will. Conscious will might activ
the progression of the voluntary process to action; it
be simply a passive observer in such a case.
We may view voluntary acts as beginning with un
initiatives being “burbled up” by the brain. The con
would then select which of these initiatives may go
an action, or which ones to veto and abort so no
appears. I discuss the full implications for free will i
section.
Robert Doty has wondered whether the unconsc
tives by the brain are so frequent that they keep the
will continuously busy with monitoring which initiativ
But we don’t know how frequently the initiatives for
acts “burble up.” They may arise relatively infrequen
case, however, unconscious processes may provide in
about the acceptability of a given initiative. These un
processes would develop during the 400 msec or so
the onset of RP II. (See the following section “Does
scious Veto Have a Preceding Unconscious Origin?”)
scious veto process need be alerted for possible a
Do We Have Free Will?
The question of free will goes to the root of our v
human nature and how we relate to the universe a
laws. Are we completely defined by the deterministi
physical laws? Are we essentially sophisticated autom
our conscious feelings and intentions tacked on as e
ena with no causal power? Or, do we have some ind
in making choices and actions, not completely dete
known physical laws?
The most common, popular view is that the hum
ual has a God-given capacity to choose or decide wh
to do, and that this capacity is not completely subje
ministic restrictions by the physical laws of nature. S
has been promoted by many of the world’s religion
this view, it becomes difficult to promote an ethics o
responsibility for one’s voluntary actions. The trad
prevalent view of free will also assumes that a person
ercised consciously. When people are completely u
their choices for action, and are performing such a
sciously, society tends to regard them as having a dim
sponsibility for their actions.
Many also believe that God is all-powerful in the
man and nature. This has produced the related belie
“fate” is ordained, and that all of one’s activities are
individual’s independence. If God knows in advance
are going to do, then, obviously, your choices for a
been made even before you yourself would have ma
140 •
Theologians have, over the centuries, devised differ
ophies to permit the existence of free will in a way th
patible with the view of the universal and all-knowing
example, a Jewish mystical sect, the Kabbalists, pro
God Himself voluntarily relinquished His power to k
vance what human beings will do. That would permit
tion of human free will, a feature that God wanted m
sess (see Cooper, 1997).
Timing of Brain Processes and Conscious Will

In our experiments, we removed all constraint on freed
tion; subjects performed a simple flick or flexion of t
any time they felt the urge or wish to do so. These
acts were to be performed capriciously, free of any ext
tations or restrictions. We have already seen that free w
be viewed as an initiator of such a freely voluntary p
clearly found that the initiation of the preparation
nate in a freely voluntary movement arises unconscio
brain, preceding the conscious awareness of wanting
ing to “act now” by about 400 msec or more.
Control Function of the Conscious Will

The existence of a veto possibility is not in doubt. Th
in our experiments at times reported that a consciou
urge to act appeared but that they suppressed or vetoe
absence of the muscle’s electrical signal when being
there was no trigger to initiate the computer’s record
RP that may have preceded the veto. Thus, there w
ous section “Conscious Veto”).
All of us, not just experimental subjects, have expe
vetoing a spontaneous urge to perform some act. Th
curs when the urge to act involves some socially un
consequence, like an urge to shout some obscenity a
fessor. Incidentally, in the disorder called Tourette’s
subjects do spontaneously shout obscenities. These
tually involuntary. No RP appears before such an ac
an RP does appear before an act voluntarily produ
Tourette patient. In any person, a quick reaction to an
stimulus also lacks a preceding RP. It is not a consc
voluntary act, although it may depend on unconsciou
prepared previously.
As noted above, in addition to a veto, there is ano
tial function for conscious will. It could serve as a tri
required to enable the volitional process to proceed
tion. That would give the conscious will a role in the
duction of the motor act. This hypothetical role fo
will has not been established experimentally. Acts th
somewhat “automatic” can be performed with no
conscious wish to do so. But the RP is rather minim
tude and duration before such automatic acts.
Obsessive-compulsive disorder (OCD) provides an
and relevant example of an abnormal relation bet
tional urges to act and the role of the veto functio
the patient experiences conscious urges to perform
repeatedly—like washing the hands over and over
clearly lacks the ability to veto each urge, and thus no
142 •
act. The patients learned to work hard to consciousl
compulsive process, and thus relieve their OCD. Sch
Begley proposed that an active “mental force” had to a
the veto of the compulsive urge to act, and that this
mental force could not be explained or accounted for
alist, determinist views. Recently, a San Francisco p
told me he has begun to train patients who have ten
act violently to veto such violent urges.
All this is in accord with my view of the conscious
tion, and it provides powerful support for my propos
free will operates. That is, free will does not initiate a
process; but it can control the outcome by actively v
volitional process and aborting the act itself, or allowi
gering) the act to occur.
In Tourette’s syndrome, a condition affecting abo
people in the United States, urges to produce vocal
with often obscene language, as well as other abn
haviors, are mostly not consciously controllable. Brai
studies (Wolf et al., 1996) have found that the caudate
involved in this malady. The caudate nucleus is one of
ganglia” located below the cerebral cortex. It appear
volved in organizing intentional movement behavio
eral. In persons with Tourette’s syndrome, the cauda
a heightened sensitivity to dopamine. By contrast,
ficiency of the neurotransmitter dopamine that is r
for Parkinson’s disease. Parkinson patients exhibit
ability to initiate a movement (among other motor
Interestingly, patients with obsessive compulsive diso
neural action on the caudate nucleus, although it wo
the veto is probably initiated in the prefrontal lobe o
bral hemisphere. As noted elsewhere, a lesion in the
lobe may result in a more uninhibited, often asocial b
In a recent book, the social psychologist Daniel We
presents a lengthy argument for the view that cons
will is an illusion. He accurately describes our e
showing that voluntary acts are initiated unconscio
brain. Wegner, like many others, states that our ex
finding indicates that conscious will “might just be a l
one of those things, like the action, that is caused by p
(p. 55) and mental events. However, nowhere in his
Wegner discuss the veto phenomenon and its provisi
tential causative role for conscious will. That role wo
of controlling the final appearance of a voluntary a
the voluntary process is initiated unconsciously befor
will appears.
Handling Our Feelings That We Initiate Volun

The view, stated in the previous section “Timing of
cesses and Conscious Will,” of how free will may op
create a problem: How can we explain our feeling or
that we initiated an act? If the cerebral process tha
freely voluntary act is an unconscious one, the feeli
sciously initiating the process becomes paradoxical.
that we do become aware of the urge (or wish) to ac
actual motor act. That could give rise to the feeling t
consciously initiated the process. However, the feelin
144 •
appears, acts as a trigger to enable the unconsciously
initiative to proceed further to production of the act
case, the conscious feeling of initiating or producing
tary act would reflect reality; it would then not be an
What we are sure of is the ability of the conscio
block or veto the volitional process and prevent the a
of any motor act. In other words, conscious free will
trol the outcome of an unconsciously initiated process
it has an additional role in enabling a nonvetoed act t
to consummation is not presently established experim
Does the Conscious Veto Have a

Preceding Unconscious Origin?
We should, at this point, consider the possibility tha
scious veto itself may have its origin in preceding un
processes, just as is the case in the development and a
of the conscious will. If the veto itself were to be in
developed unconsciously, the choice to veto would the
an unconscious choice of which we become conscio
than a consciously causal event. Our own previou
showed that the brain “produces” an awareness of
only after about a 0.5-sec period of appropriate neuro
tions (see Chapter 2, and reviews by Libet, 1993, 1996)
Some have proposed that even an unconscious init
veto choice would nevertheless be a genuine choice m
individual that could still be viewed as a free will proc
ample, Velmans, 1991). I find such a proposed view of
be unacceptable. In such a view, the individual woul
processes. But, a free will process implies one can b
sciously responsible for one’s choice to act or not
do not hold people responsible for actions perform
sciously, without the possibility of their conscious co
For example, actions by a person during a psycho
leptic seizure, or by one with Tourette’s syndrome (s
cially abhorrent epithets), are not regarded as actio
will. Why then should an event unconsciously deve
normal individual, if it were a process over which he
conscious control, be regarded as an act of free will fo
should be held responsible?
I propose, instead, that the conscious veto may not
be the direct result of, preceding unconscious proc
conscious veto is a control function, different from
coming aware of the wish to act. There is no logi
tive in any mind-brain theory, even in identity t
requires specific neural activity to precede and determ
ture of a conscious control function. And there is no
tal evidence against the possibility that the control p
appear without specific development by prior uncon
cesses.
Admittedly, to be conscious of the decision to veto
one is aware of the event. How may one reconcile th
proposal? Perhaps we should revisit the concept of
and how it relates to the content of awareness, in ce
cesses that develop both awareness and its contents
previous studies have indicated that awareness is a u
nomenon in itself, distinguished from the contents of
may become aware.
146 •
But the content of those awarenesses in these two cas
ent; with the cortical stimulus the sensory awareness
tively delayed, while it is not subjectively delayed with
lus to the subcortical pathway. The content of an un
mental process (for example, correct detection of a s
out any awareness of the signal) may be the same as t
(correct detection) when there is awareness of the sig
become aware of that same content required that th
of stimulus to the subcortical pathway be increased
400 msec! (See Libet et al., 1991.)
In an endogenous, freely voluntary act, awareness
tention to act is delayed for about 400 msec after brain
initiate the process unconsciously (see the previous se
quence of Events in the ‘Act Now’ Situation”). Aware
oped here may be thought of as applying to the whole
process. That includes the content of the conscious u
and the content of factors that may affect a consc
Awareness of an event may not necessarily be restric
detailed item of content in the whole event.
The possibility is not excluded that factors on whic
sion to veto is based do develop by unconscious pro
precede the veto. However, the conscious decision to
still be made without direct specification for that decis
preceding unconscious processes. That is, one could c
accept or reject the program offered up by the whol
preceding unconscious brain processes. The awaren
decision to veto could require preceding unconscious
but the content of that awareness (the actual decision
a separate feature that need not have the same require
Can we assume that voluntary acts, other than the
studied by us, also have the same temporal relation
unconscious brain processes and the appearance o
scious wish or will to act? It is common in scientific
be limited technically to studying a process in a sim
and then to find that the fundamental behavior disco
the simple system does indeed represent a phenomen
pears in other related and more complicated systems
ple, the charge on a single electron was measured by
one isolated system, but it is also valid for electron
tems. In fact, RPs have been found by other investiga
cede other more complex volitional acts, such as be
speak or to write. Those investigators did not, how
the time of appearance of the conscious wish to begi
We may, therefore, allow ourselves to consider what
plications may follow from our experimental find
viewed as a feature of voluntary acts in general.
We should also distinguish between deliberations
choice of action to adopt (including preplanning of w
on such a choice), and the final intention to actually
One may, after all, deliberate all day about a choice
act. There is no voluntary act in that case. In our ex
studies we found that in some trials subjects engag
conscious preplanning of roughly when to act (say,
second or so). But even in those cases, the subject
times of the conscious wish to “act now” was about
This value was very close to the values reported for fu
148 •
These findings indicate that the sequence of the cer
tional processes to act now may apply to all volit
whether they are fully spontaneous or have a histo
scious deliberations. That is, regardless of the prese
sence of advance deliberation or planning, the process
now arises unconsciously, about 400 msec before the
wish to act now appears. The “act now” process app
independent of and separate from the deliberative an
processes.
Ethical Implications of How Free Will Operates

The role of conscious free will would be, then, not t
voluntary process (although it may possibly enable t
to finally lead to action). However, conscious will defi
control whether the act takes place. We may view t
scious initiatives for voluntary actions as “burbling u
sciously in the brain. The conscious will then selects
these initiatives may go forward to an action, or whi
veto and abort so no act occurs.
This kind of role for free will is actually in accord
monly held religious and ethical strictures. Most rel
losophies hold individuals responsible for their ac
advocate that you “control your actions.” Most of the
mandments are “do not” orders. The philosopher an
sage Maimonides “defined holiness as disciplined self-
the ability to say no to one’s most instinctive physical d
quoted in Rabbi Shlomo Riskin, 1999). In this connecti
an interesting difference between the Jewish and Chr
people alone, with tolerance. The Christian view ta
tive, activist view: “Do unto others what you would
do unto you.” The late philosopher Walter Kaufma
that this difference is highly significant in his book Fa
etic (1961); Kaufmann noted that, among other things
tian Golden Rule could result in actions being impo
ers that run counter to the wishes of others.
When May One Be Guilty or Sinful?

How do our findings relate to the question of when o
regarded as guilty or sinful, in various religious and
cal systems? If we experience a conscious wish or u
form a socially unacceptable act, should that be reg
sinful event, even if the urge has been vetoed and no
curred? Some religious systems answer “yes.” Presid
Carter admitted to having “a lust in his heart” for som
Although he did not act on this, he apparently still fe
having experienced a lustful urge. (President Carter w
on a Christian tradition deriving from the following
in the Sermon on the Mount: [Jesus said], “ye have h
was said by them of old time, Thou shalt not comm
But I say unto you, that whosoever looketh on a wom
after her hath committed adultery with her already in
(Matthew 5:27–28, recalled for me by Rev. Anthony F
But any such urges would be initiated and develo
brain unconsciously, according to our findings. The u
appearance of an intention to act could not be cont
sciously. Only its final consummation in a motor ac
150 •
out, would create a physiologically insurmountable
psychological difficulty.
Indeed, insistence on regarding an unacceptable urg
sinful, even when no act ensues, would make virtua
viduals sinners. The mechanism for the unconscious
of such an urge is presumably present in all human b
all persons are undoubtedly going to experience soc
ceptable urges and intentions to act. In that sense su
could provide a physiological basis for original sin! O
the concept of original sin could also be based on diffe
about what is regarded as sinful.
Ethical systems deal with moral codes or conven
govern how one behaves toward or interacts with oth
uals. They are presumably dealing with actions, not si
urges or intentions. Only an action by one person c
impinge on the welfare of another. Because it is the p
of an act that can be consciously controlled, it should
mate to hold individuals guilty of and responsible for
Determinism and Free Will

There remains a deeper question about free will that t
ing considerations have not addressed. What we hav
experimentally is some knowledge of how free will
ate. But we have not answered the question of (1) w
consciously willed acts are fully determined by natura
govern the activities of nerve cells in the brain, or (2
freely voluntary acts, and the conscious decisions t
them, can proceed to some degree independently of n
brain’s activities with no causal powers of its own.
The view that free will is illusory is elaborated at s
by Wegner (2002). There are, of course, other cont
this view, like the Churchlands (1999) and Denn
Wegner proposes a “theory of apparent mental caus
states: “People experience conscious will when the
their own thought as the cause of their action” (p
book). That is, the experience of conscious will is “qui
dent of any actual causal connection between the
and their actions.” It is, of course, legitimate to prop
rangement as a theory for free will within a determi
But there is no crucial evidence that proves its validity
imental test has even been proposed in which this th
be falsified. Without any possibility of falsification, o
pose anything without any fear of being contradict
Popper explained).
First, free choices or acts are not predictable, even
viewed as completely determined. The “uncertainty
of Heisenberg precludes our having a complete kno
the underlying molecular activities. Quantum mecha
us to deal with probabilities, rather than with cer
events. And, in chaos theory, a random event may s
havior of a whole system in a way that was not p
However, even if events are not predictable in practic
not exclude the possibility that they are following n
and therefore determined.
Let us rephrase our basic question as follows: Mus
determinism? Is nondeterminism a viable option?
152 •
Determinism (adherence to natural law) has, on t
worked well for the physical observable world. Th
many scientists and philosophers to regard any devia
such determinism as absurd, witless, and unworthy o
ation. But natural laws were derived from observation
cal objects, not from subjective mental phenomena.
cannot be directly observed; they are inner experience
dividual who has them. There has been no evidence,
proposed experimental test design, that definitively o
ingly demonstrates the validity of natural law deter
the mediator or instrument of free choice or free will
There is an unexplained gap between the categor
cal phenomena, and the category of subjective pheno
searchers as far back as Leibniz have pointed out t
looked into the brain with a full knowledge of it
makeup and nerve cell activities, you would see no
described subjective experience. You would only s
structures, their interconnections, and the production
impulses and other electrophysiological events, as we
bolic chemical changes. The foundation of our own e
tal studies of the physiology of conscious experience
in the late 1950s) was that externally observable brain
and the related reportable subjective introspective e
must be studied simultaneously, as independent cat
understand their relationship. The assumption that a d
tic nature of the physically observable world can a
subjective conscious functions and events is a spec
lief, not a scientifically proven proposition. (Of cours
exclude the possibility that physical events are suscep
external “mental force” at the micro level, in a way
not be observable or detectable).
Nondeterminism—which is the view that consciou
at times, exert effects not in accord with known phys
is of course also a nonproven speculative belief. Th
conscious will can affect brain function in violation
physical laws takes two forms. One view is that the
are not detectable, because the actions of the mind m
level below that of the uncertainty allowed by qu
chanics. (Whether this last proviso can in fact be t
matter yet to be resolved.) This view would thus
nondeterministic free will to occur without a perce
tion of physical laws. A second view holds that vi
known physical laws are large enough to be detecta
in principle. But it can be argued that detectability in
tice may be impossible. That difficulty for detection w
pecially true if the conscious will is able to exert its in
minimal actions at relatively few nerve elements, if th
could serve as triggers for amplified nerve cell patter
ity in the brain. In any case, we do not have a scientifi
the question of which theory (determinism or nonde
correctly describes the nature of free will.
However, it is important to recognize an almost u
perience: that we can act in certain situations with
dependent choice and control of whether to act. T
example of this is the one we employed in our ex
study—the conscious will to flex the wrist in a freely
154 •
ence must be qualified. Our own experimental findin
that conscious free will does not initiate the final “act
cess; the initiation of it occurs unconsciously. But, as
previously, conscious will certainly has the potential
trol the progress and outcome of the volitional proc
the experience of independent choice and of co
whether and when to act) does have a potentially so
as not being an illusion. The cerebral nature of c
choices of action, by conscious deliberation and prepl
fore any “act now” process, is yet to be elucidated.
How does this experience dovetail with the view of
mental scientist? It appears to create more difficulty f
minist than for a nondeterminist option. The phenom
that most of us feel that we do have a kind of free w
for some of our actions, within certain limits that m
posed by our brain’s status and by our environment.
itive feelings about the phenomenon of free will form
mental basis for our views about human nature. G
should be taken not to believe allegedly scientific c
about our nature that depend on hidden ad hoc assum
theory that simply interprets the phenomenon of free
lusory and denies the validity of this phenomenal fac
tractive than a theory that accepts or accommodate
nomenal fact.
Given that the issue is so fundamentally importa
view of who we are, a claim that our free will is illuso
be based on fairly direct evidence. Theories are suppo
plain observations, not do away with them or distort
that free will is illusory, like that of Wegner (2002), f
category. It is foolish to give up our view of ourselve
some freedom of action and of not being predeterm
on the basis of an unproven theory of determinism.
My conclusion about free will, one genuinely free
determined sense, is that its existence is at least as go
better, scientific option than is its denial by natural
minist theory. Given the speculative nature of both
and nondeterminist theories, why not adopt the view
have free will (until some real contradictory evidence
it ever does)? Such a view would at least allow us to p
way that accepts and accommodates our own deep
we do have free will. We would not need to view o
machines that act in a manner completely controlled
physical laws. Such a permissive option has also be
advocated by the neurobiologist Roger Sperry (see D
I close, then, with a quotation from the great no
Bashevis Singer. Singer stated his strong belief in our
will. In an interview (Singer, 1968), he volunteered
greatest gift which humanity has received is free c
true that we are limited in our use of free choice. B
free choice we have is such a great gift and is potent
so much that for this itself life is worthwhile living.”
156 •
5
CONSCIOUS MENTAL FIELD THEORY: EXPLAININ
THE MENTAL ARISES FROM THE PHYSICAL
Present day physics represents a limiting case—valid

inanimate objects. It will have to be replaced by
laws, based on new concepts, if organisms with
sciousness are to be described.
—Eugene Wigner, Nobel Laureate, Physics (Quo

Burns, 1991)
What Is the Problem?

Perhaps the most profound question we can ask is,
conscious subjective experience arise from activities
cells in the brain? That is, how can the mental arise
physical? There is no doubt that appropriate neural a
the human brain are essential for the mental (subject
ence) to appear. Yet, if you were to look into this neu
and the neuronal structures involved, you would no
thing that looked like subjective experience. Is it pos
tack this profound question experimentally? To atte
to the individual having the experience. Yet it appears
lation to appropriate neural activities in the brain. To
quires an introspective report by an individual descri
perience or awareness of something.
We have seen that the relationship between consci
ence and neuronal activities in the brain can be studi
fully by examining the two features together, as indep
interrelated variables. That is, the neuronal activiti
conscious experience must be observed for the sam
order to discover which brain activities may signify a
experience. Any significance of these for conscious
would only be established by concomitant studies o
experience with brain functions.
However, even a successful investigation of the
relationship between conscious experiences and neur
ties, important as that is, will not answer a more prof
lem: How does the categorically different nonphysica
non of subjective experience come from the physic
of nerve cells? This problem has been termed the “
lem” by philosopher David Chalmers (1996).
Chalmers (1995) tried to solve this problem by p
double-aspect theory of information. He postulated i
as having both physical and phenomenal aspects.
would emerge from or be identical with the phenome
This proposal, which appears to be a version of iden
is unconvincing for various reasons (Libet, 1996). Iden
posits that there is a common “substrate” for all real
158 •
C O N S C I O U S M E N TA L F I E L D T H E O R Y
sible to an external observer. But identity theory, inc
Chalmers’s version of it, is not testable; it is therefore
entific theory. A different testable solution, the unified
mental field, is presented in this chapter.
Philosopher Colin McGinn (1999) views this prob
“unanswerable question.” He sees no conceivable wa
we could account for consciousness and subjective exp
terms of the natural physical order. Spinoza believed
the problem by arguing that “thoughts and experien
ways identical with a sequence of objective changes in
body.” This view suffers from the difficulties with id
ory in general. For one thing, it is not testable, and i
appear to offer an explanation of how the mental and
cal are causally interactive; it rests on a metaphysical b
if an attractive one. I shall attempt to show that it is
propose a testable theory as an answer to the probl
mental and physical.
In addition to Colin McGinn and David Chalme
other philosophers are prominent in their writings on
lem of how to relate brain activities to conscious
experience. These include John Searle (1992) and D
nett (1991) along with Paul and Patricia Churchland (
Churchlands represent one extreme in the range of v
mental subjective phenomena are reducible to physica
nerve cells. They “advocate a doctrine of eliminative
ism. The view, put baldly, is that mental states do no
talk as if they do when we use what has come to be c
C O N S C I O U S M E N TA L F I E L D
On the other hand, Searle (1992) views conscious
as a real phenomenon, one that is not reducible to t
activities of the neurons in the brain. That view is,
not unique to Searle. My own experimental work sin
1950s is based on the irreducibility of the two catego
nomena, the mental and the physical, to each othe
also shares this view, though, like me, he is not in
with Searle’s further elaborations of this view.
If the Churchlands want to regard themselves as
determined by the physically materialistic events of
they are entitled to their view, even though others fee
a real conscious mind that is not an automaton. Ind
Descartes in the sixteenth century who asked, What
sure of being real? His answer was, It is only my own
subjective mind or experience that I am sure is real.
Searle claims that achievement of a complete kno
the neural events that are correlated with a conscious
will tell us all we can know about the mind-brain r
Searle views consciousness as simply a biological high
cess of neurons in the brain. McGinn (1999) poin
Searle does not answer the central question of ho
logical process of conscious experience results from
physical properties of neurons. What “Searle offers a
to the problem is really just a statement of it.”
Searle (2000a) went further to present a view of
scious experience and free action relate to brain
Searle’s model for this view is not in accord with the
160 •
findings show that the process to “act now” is initiat
sciously. Searle’s model proposes that “freedom of th
pears during a “gap” between making a decision to a
onset of the action process. But the gap is an uncon
of about 400 msec between initiation of the decisio
conscious decision. (See further in my commentary o
proposal, Libet, 2001.) As is the case with many of th
phers’ speculative views on the mind-brain question
models have not been tested and are not even testa
mentally.
The unity of conscious experience is a phenomenon t
of the “hard problem.” An obvious example lies in o
ence of a visual image. The image is based on the a
thousands of nerve cells. The spatial pattern of these
in the primary receiving visual area of cerebral cortex
look like the image we see; it is distorted. Other f
the visual image are represented or developed in oth
visual areas; these are specialized for color vision,
an image, recognition of faces, and so on. In spite of
plex array of separable neural functions, we see sub
unified image, with all of these visual elements sm
tegrated. This is only one small example of our un
grated experience arising from complex patterns of
activities.
Still another feature of the hard problem is the q
how free will operates. If free will is accepted as a valid
enon, a conscious mental function (presumably nonp
Indeed, “five mysteries of the mind” have been el
forth by Doty (1998).
How Do Emergent Phenomena Relate to the Pr

In the physical world we recognize that the phen
played by a system may not be evident in the prope
subunits that make up that system. For example, the
benzene is composed of six carbon and six hydro
Kekulé proposed that the six carbon atoms are con
ring structure, with the hydrogens bound to the rin
junctions between the six carbon atoms in the ring.
famous benzene ring fundamental to organic chemis
biology). The properties exhibited by benzene (as an
vent and so on) could not have been predicted, a prio
properties of the carbon and hydrogen atoms thems
is, a new property emerged from the C6H6 ring system
the properties of a wheel are not evident from the pr
the materials used to construct the wheel (as noted
1980). The wheel’s property of rolling emerges from
created by the specific arrangement of the materi
make the wheel. The appearance of a magnetic fiel
wire that is conducting an electric current is also a ph
that emerges in that system. As the astrophysicist Sir
dington once said, “We used to believe that if we
thing and then another, then we know two, becau
one are two. We are now discovering that we must
about the ‘and’” (quoted by B. D. Josephson, 1993
162 •
ence similarly, as a phenomenon that somehow eme
an appropriate system of activities of the physical n
in the brain. However, unlike physical emergent ph
the emergent subjective experience is not directly obs
measurable by any physical means, as subjective exp
only accessible to the individual having the experienc
the emergent subjective experience of this system is
properties of the responsible nerve cells; it could not
a predicted outcome of these neural activities. It s
be surprising that the emergent subjective experienc
unique unexpected characteristics.
Why subjective experience emerges from approp
ronal activities may be no more answerable than sim
tions about other fundamental phenomena. That is,
mass have inertia? Why do masses exhibit gravitatio
tion? Why does matter behave both in wave-like an
fashions? Fundamental physical phenomena are not re
explainable. We simply accept these as “given” in the
things. We can only study how these manifestations af
act, and control what goes on in the physical world.
We may thus regard conscious, subjective experie
other unique fundamental property in nature. What
of its unique characteristics, in addition to subjective
or awareness itself ? These are the unity of subjective
and the potentiality for influencing nerve cell activit
characteristics are also not evident in the neuronal
from which subjective experience emerges.
One of the most mysterious and seemingly intractabl
in the mind-brain relationship is that of the unitar
grated nature of conscious experience. We have a br
estimated 100 billion neurons, each of which may
sands of interconnections with other neurons. It is i
evident that many functions of cerebral cortex are
This is not merely true of the primary sensory are
sensory modality, of the motor areas that command
and of the speech and language areas—all of which
known for some time. Many other functions now fin
representations, including visual interpretations of co
and velocity of images; recognition of human faces;
ration for motor actions. Localized function appear
even to the microscopic level within any given area
bral cortex appears to be organized into functional an
cal vertical columns of cells, each a millimeter or s
There are discrete interconnections within the colum
other columns near and far, as well as with selective s
structures. This columnar view began with findings
castle (1957) and has been greatly extended by him
ers. For example, there are the columnar localization
shapes and motions and of binocular vision, as dis
Hubel and Wiesel (1962).
In spite of the enormously complex array of loca
tions and representations, the conscious experiences
or elicited by these neuronal features have an integrat
fied nature. For example, we subjectively see a smoo
164 •
as an infinitely detailed array of widely individual eve
be argued that this amazing discrepancy between part
neuronal representations and unitary integrated consc
rience should simply be accepted as part of a gener
isomorphism between mental and neural events. Bu
applies some verbal names to the phenomenon. It
profound question of how that mismatch leads to unifi
ence. One would not like to exclude the possibility
unifying process or phenomenon may mediate the
transformation in question.
The general problem of subjective unity in the fac
cal disunity has been recognized by many others, g
at least to a founder of modern neurophysiology, S
(1940), and probably earlier. Eccles (in Popper and E
p. 362) proposed that “the experienced unity comes n
neurophysiological synthesis but from the proposed i
character of the self-conscious mind.” He made this p
conjunction with a dualist-interactionist view in whi
rate nonmaterial mind could detect and integrate the
activities. Some more monistically inclined neuroscie
also been arriving at related views (for example, Sp
1980; Doty 1984)—in other words, integration seems
accountable for in the mental sphere that emerges
neuronal activities.
There has been a growing consensus that no sin
group of cells is likely to be the site of a conscious e
but rather that conscious experience is an attribute
global or distributed function of the brain (for exa
give rise to a conscious experience (Libet, 1973, 198
al., 1991). For example, a short train (say of 100-msec d
stimulus pulses to sensory cortex elicits response
nerve cells without any subjective experience.
More recently, a widespread synchronization of
neuronal responses to certain visual configurations
ered (Gray and Singer, 1989; Singer, 1991, 1993). Singer
that their results “provide experimental support fo
postulate of Edelman’s group selection theory.” Th
led to some speculation that a “correlation” model m
sent the neural coding for recognizing a unified imag
erwise chaotic background. That is, the synchronous
of electrical oscillations would give a unified subjec
This speculation is still to be tested directly. But even
correlation between synchronization of neurons an
subjective experience were to be found, that would
why the subjective experience is unified in a comple
and with no gaps in the spatial and colored image,
synchronized activities of separate groups of nerve c
How Does Free Will Arise?

Another apparently intractable problem in the mind
tionship is the question of whether the interaction b
mind and the brain can go in both directions. There
that cerebral events or processes can influence, co
presumably “produce” mental events, including cons
The reverse of this principle, that mental processes
ence or control neuronal ones, has been generally un
166 •
provide prima facie evidence for such a reverse interac
This reverse feature is obviously fundamental to th
free will (see Chapter 4). There have been many view
historically about mind influencing brain, mostly f
logians and philosophers. These have had importan
vocative effects on the general populations of the wo
ever, they are virtually all not testable by objective o
criteria.
Even the serious and detailed theses proposed
scientists have been speculative solutions that are
provoking but are not experimentally testable. Sir Jo
(1990), a Nobel laureate in neurobiology, proposed a d
lution. He proposed that mental units (called psychons
rate from nerve cells but can affect the probability of
the chemical transmitter substance at synaptic junctio
power could then influence the ability of a given ne
deliver a message to the next cell in its network. Ro
(1980), the Nobel laureate who established that the rig
sides of the brain could function differently and eve
dently, argued for a monistic solution that did not se
mental and physical attributes of brain function. S
posed that mental activity emerges from the physic
the brain. But the emergent mental activity could, in t
ence neuronal activity in the brain. He limited that in
“supervening” but not “intervening” in neuronal act
limitation allowed Sperry to remain a determinist in
However, after struggling for decades with the proble
to accommodate a humanistic free-will aspect of hum
pletely governed by the natural laws of the physical
Doty, 1998). Unfortunately, both these views (of
Sperry) remain philosophical theories having explana
but without experimentally testable formats.
Does a Unified Conscious

Mental Field Provide a Solution?
As one possible experimentally testable solution to bo
of the mind-brain relationship, I have proposed th
view conscious subjective experience as if it were a
duced by appropriate though multifarious neuronal a
the brain (Libet, 1993, 1994). Such a field would provi
nication within the cerebral cortex without the neu
tions and pathways in the cortex.
A conscious mental field (CMF) would provide th
between the physical activities of nerve cells and
gence of subjective experience. It thus offers an ans
profound question of the nonphysical mental arisin
physical.
A chief quality or attribute of CMF would be that o
or unitary subjective experience. That is, the CMF w
entity in which unified subjective experience is pres
ond attribute would be a causal ability to affect or
neuronal functions. The additional meaning or e
power of describing subjective experience in terms
will become more evident with the proposed experim
ing of the theory. That is, the CMF is proposed as mo
another term for referring to “unified subjective expe
168 •
independent category; it is not describable in terms of
nally observable physical events or of any known ph
ory as presently constituted. In the same sense as for
tive events, the CMF would be detectable only in
subjective experience, accessible only to the individua
the experience. An external observer could only gain v
evidence about the conscious mental field from an int
report by the individual subject. In this respect the
mental field would differ from all known physical fie
existence and characteristics are amenable to physica
tions. The CMF theory may be viewed as an extension
Sperry’s theory of “the mental” as an emergent prope
physical” brain.
The proposed CMF should be viewed as an operat
nomenon, in other words, as a working and testable
brain function. You may think of the CMF as somew
gous to known physical force fields (Libet, 1997, follo
per et al., 1993). For example, a magnetic field is pr
electric current flowing in a conductor, but it can in
ence the flow of the current. However, as indicated
cannot be observed directly by external physical mean
How is the CMF attribute of unified subjective exp
lated to its production by contributions from loca
areas? Local alterations in the CMF would be refl
changed overall field, but there would not be a sep
quired mechanism for transmission and integration of
contributions. To think in terms of a transmission an
tive process would be to continue thinking in terms
ternally observable neural events. To do so would b
rules for (at least much of ) the relationship betwee
and the physically (externally) observable neural pro
the rules are not describable a priori—in other wo
they are discovered by studying both phenomen
neously (see Libet, 1987, 1989).
In the split-brain studies of Sperry et al. (1969; Sp
the main communicating commissures, the large
nerve fibers connecting the two cerebral hemisph
transected, or cut through. Neurosurgeons do that
epileptic seizures that bounce back and forth betwe
cerebral hemispheres. The researchers then found
can be simultaneously different contents of experie
two sides. Normally the two hemispheres talk to eac
means of the large commissures and they share the
mation. With the commissures split, however, the ne
of mental events in the right hemisphere are not avai
left hemisphere, and vice versa. As a result, any contr
right hemisphere activity to a mental field presumably
rectly alter a CMF of the left hemisphere. Unity o
would, in these circumstances, be restricted to a g
sphere. In addition, contributions of local neural a
overall CMF of a hemisphere would only be effective
tiguous with those of other areas; in other words, th
tions would not be effective across substantial gaps
of tissue barriers, of the dimensions present betwe
hemispheres. If the CMF cannot cross a barrier betwe
adjacent hemispheres, it clearly cannot provide a bas
sages to be transmitted to or received from anoth
170 •
persons, if one hemisphere cannot communicate wit
cent hemisphere in the absence of the major interc
nerve bundles (see also Buser, 1998).
Incidentally, these features raise other fundamental
Is the right hemisphere conscious? Are there two sel
individual, one self in each hemisphere?
The right hemisphere does appear to be capable of
scious, even though it has a very limited ability to spea
opportunity to see a video of a patient from who
hemisphere had been surgically removed due to its
The excision was done when the patient was an adul
viated the possible adjustments that can occur in the r
sphere when the left hemisphere is removed or otherw
in childhood. This adult patient behaved as if he were
He looked alert; he responded to questions in a prope
At times, because he was not able to answer by sp
showed he was frustrated and disgusted by this limita
The question of two selves is a more complicate
Bogen, 1986; Doty, 1999). Split-brain patients do not
turbances of their feelings as a unified person. That i
like the same single person they were before the sur
When not being tested with inputs restricted to one he
both hemispheres can receive the same sensory in
These patients’ eyes can roam about over all the sa
fields. Still, it is quite remarkable that they do not rep
that there is a partner conscious agent; they feel th
one self.
We may postulate that some aspect of the putative
Is There an Experimental
Design to Test the CMF Theory?
Any scientific theory, especially one like the CMF, m
able to be taken seriously. The theory of a CMF ma
predictions that can, at least in principle, be tested
tally. If local areas of cerebral cortex can independen
ute to or alter the larger, unitary CMF, it should be
demonstrate such contributions when (1) the cort
completely isolated or cut off from all neuronal com
with the rest of the brain; but (2) the area remains i
and kept functioning in some suitable manner that
resembles its normal behavior. The experimental pr
be tested would be as follows: Suitable electrical or c
tivation of the isolated tissue should produce or af
scious experience even though the tissue has no neu
tions to the rest of the brain. Communication would
to take place in the form of some field that does not
nerve pathways.
The researchers would have to control for the pos
spread of influences from the isolated block via ph
neural paths (for example, electric current flow). If a
experience were induced and reported within a sec
that would tend to exclude spread by chemical diffus
as by changes in vascular circulation or in contents of
blood (see Ingvar, 1955).
Suitable neuronal isolation could be achieved ei
172 •
of an area. Surgical isolation is discussed further la
chapter.
Functional isolation might be achievable by injectin
agents in small amounts to form a ring of blockade a
under a selected block of cerebral cortex. A local
agent might be used, such as procaine suitably buffe
7.4 in Ringer’s solution. Or tetrodotoxin, the selective
sodium-conducted action potentials, could be combin
calcium-channel blocker like verapamil (to ensure tha
mediated action potentials did not escape blockade). T
tage of this pharmacological method for isolation is
ibility, which means it could be used on areas of
scheduled for surgical excision, thus greatly enlargin
of potential subjects (if risk factors are suitably met).
vantages of this method are (1) the difficulty of lim
blockade to a narrow band around the slab, because
bility; (2) the need to prove that complete blockade
achieved; and (3) a reduced ability to introduce neu
into the isolated slab, by the excitation of ascending n
within the slab but near its lower borders. The chem
ade would inactivate some portion of these nerve fibe
diffusion.
How Can a Surgically Isolated Slab

of Cerebral Cortex Be Produced in situ?
A slab of cerebral cortex can be neurally isolated sur
making all of the cuts subpially, the blood supply to th
contact with the surface of the brain, including the ce
tex. The blood vessels to the cortex travel horizontally
membrane. At separate points a branch of blood v
down vertically into the cortex. Cuts of the cortex ca
just below the pia, leaving the blood vessels intact.
Studies of the electrophysiological activity of su
cortex in situ have been reported (Kristiansen and
1949; Burns, 1951, 1954; Echlin et al., 1952; Ingvar, 195
et al., 1961). The basic method involved introducing
curved blade through an opening in an avascular a
pia arachnoid membrane. The surgeon would under
or slab of cortex and, by bringing the tip of the cu
up to meet the pia at some distance away and rota
in a circle, he would also cut the connections to adjac
In an earlier study (of how vertical cuts in the conn
tween adjacent cortical areas might affect the integr
nized function of the sensorimotor cortex in monke
(1947) used a somewhat different technique (Fig. 5.1
ting instrument was an extremely thin double-edged
from a fine wire or sewing needle. The sharpened e
of this wire was bent to a right angle; this terminal
the blade was sunk vertically into the cortex so its
arm lay just below the pia. When the vertical knife w
forward it cut through the cortex while its horizont
arm slid just below the pia. This technique could e
ranged to produce undercutting of the cortex as we
vantage of Sperry’s method lies in the very thin lin
damage created by this knife, which produces chron
174 •
Fig. 5.1. Production of a viable isolated slab of cerebral cortex.

A. A fine wire is bent into the shape as shown. The leading ver
is sharpened before insertion into the cortex. The vertical leadin
inserted into the cortex to a depth at which its horizontal arm i
low the surface (pia arachnoid) membrane. Pushing the vertical
ward now cuts off that side of the cortex from adjacent connec
cutting all sides and then arranging the knife to undercut the
piece of cortex becomes completely disconnected from all neura
tions with the rest of the brain, but it remains alive because it r
blood supply from vessels in the surface membrane.
B. Shows a related shape of the knife used to cut down into th
gyrus that borders the sulcus (space) next to the adjacent gyrus.
From Sperry, 1947. With permission from the American Phy
Society.
than 100 µm thick. That is particularly desirable if th

plans to leave the isolated slab in situ for therapeut
When a piece of cortex must be excised to remove
table epileptic focus, simply isolating the offending f
in the manner described here appears advantageous
proach avoids the growth of dangerous scar tissue th
occur in a cavity produced by removal of the focal co
with both general and local anesthesia (patient awake
ported an immediate reduction but not complete a
rhythmic electrical activity (EEG) in the isolated are
minutes, paroxysmal bursts of high-voltage activity
This kind of seizure pattern in a normal brain is usu
ated with disruption or distortion of normal functi
the motor area, convulsive motor actions. There wa
of activity from the isolated slab to surrounding area
The physiological properties of the cortical slab ar
immediately altered when it is isolated, because of
loss of all neural inputs. For example, it is well know
struction of the reticular activating system in the b
which ascends to end diffusely in the cerebral cortex,
coma. This afferent input would have to be properly
as to “wake up” the isolated slab of cortex. Proced
store some levels of sufficiently normal activity woul
sary. The nerve fiber inputs from below, and their sy
tacts with cells in the slab, would degenerate some
the cuts separated them from their cells of origin to p
lation. Therefore the studies proposed in my experi
sign (to test for the CMF) need to be carried out in
phase, during the initial period after isolation. With
cut axons still viable and potentially functional, the
used to restore some degree of neural inputs to the
in the slab. These ascending nerve fibers could be
excited by the surgeon inserting fine stimulating el
reach the lower levels of the isolated slab. Electri
ings of activity at the surface of the slab could serv
176 •
How to find the proper patient and tissue? Surgical iso
slab of cortical tissue would, of course, result in a p
loss of its normal neuronal functions. As a result, s
limited to patients in whom a slab of cortical tissue
therapeutically designated for surgical removal. The
would be carried out in the operating room before th
of the tissue, if other conditions were also met. T
would have to be awake and responsive; the surgeon w
to use local rather than general anesthesia to produce
cal exposure; the patient would have to provide info
sent and be cooperative; and risk assessments would
approved by all concerned, particularly the hospital
sity committee for protection of human subjects. Man
have tolerated brain surgery under local anesthesia and
ticipated fruitfully in past studies (for example, Pen
Libet et al., 1964; Libet, 1973). In such procedures, it i
for the neurosurgeon to include a bit of fairly norma
ing tissue within the slab scheduled for excision; luck
surgeons almost always include such normal bits to
adequately therapeutic removal of pathological tissue
There is a further requirement for this experiment
stimulation of the cerebral cortex elicits a reportable
experience only at very limited sites. The most effectiv
sites are the primary sensory areas, which receive the
localized sensory messages. There is a primary senso
body sensations (in the postcentral gyrus of cortex),
(in the so-called striate cortex of the occipital lobe), an
tion (in the upper lip of the temporal lobe). Stimul
the cortex, the researcher had better do that on an ar
such stimulation normally elicits an introspective re
experience.
Very few patients—five to ten worldwide each y
this requirement. Even when patients have an epilepti
primary sensory area, surgeons are very reluctant to c
focus, as the patients then may suffer a serious loss o
To find a cooperative neurosurgeon with access to su
has proved to be a daunting process.
Can the CMF Influence Nerve Cell Activity?

A test of the causal ability of the putative CMF to
ronal functions is already implicit in the test just de
the existence of the CMF. If stimulation of the isola
slab can elicit an introspective report by the subjec
must be able to activate the appropriate cerebral are
to produce the verbal report.
The proposed experiment, using an isolated but liv
cerebral cortex, could provide a direct answer to the
whether a CMF could affect nerve cell functions in
accounts for the role of conscious will.
Other proposals for the possible actions of the con
are subject to ambiguity in interpretation. For exam
activity (as indicated by measurements of regional blo
metabolic rate) has been shown to increase selecti
supplementary motor area (SMA) when a subject
imagine moving her fingers without actually moving
178 •
difficulties with such a conclusion from that experime
dications of an increase in neuronal activity by th
methods, whether PET scan or MRI, are based on an
local blood flow or metabolism. But such increases in
or metabolism appear only after a substantial delay
seconds, following any actual change in local nerve c
The delay obscures the determination of the relative
the mental imaging and the increase in local nerve c
Additionally, there is always the possibility that the w
cess was initiated by some neural events elsewhere in
too small or oriented so they are not recorded by th
method. Unless the mental event (of imagining or c
could be shown to precede any possible neural event
related to the process studied, there would always
about the nature of the causal interaction. With the ne
lated cortical slab, there are no such difficulties of inte
General Conclusions on CMF Theory

Suppose that the experimental results prove to be p
other words, suitable stimulation of the neurally isola
elicits some reportable subjective response that is no
able to stimulation of adjacent nonisolated cortex o
cerebral structures. That would mean that activation
cal area can contribute to overall unified conscious
by some mode other than by neural messages de
nerve conduction. This result would provide crucial s
the proposed field theory, in which a cortical area can
With such a finding, you might ask, What would
of all the massive and complex neural interconnectio
cortical, cortical-subcortical, and hemisphere to h
And here is a possible answer: to subserve all the cer
tions other than those directly related to the appeara
conscious subjective experience and its role in consc
should be noted that all cognitive functions (receip
recognition of signals), information storage, learning
ory, processes of arousal and attention and of state
and mood, and so on) are not proposed as functions
nized or mediated by the postulated CMF (conscio
field). In short, it is only the phenomenon of conscious s
perience, associated with all the complex cerebral fun
is modeled in the CMF, in an admittedly speculative m
Some may easily dismiss the prospect of obtaining
sults in the proposed experimental tests, because s
would be completely unexpected from prevalent view
functions based on physical connectivities and intera
the improbability of positive results is strictly a functi
ing views that do not deal successfully with the p
unity of subjective experience and of apparent men
of brain processes. The potential implications of the
ory and of the positive results it predicts are clearly p
nature. On those grounds, and because the propo
ments are in principle workable though difficult, th
experimental design should merit a serious place in
tions of the mind-brain problem.
As the theoretical physicist Niels Bohr once said a
180 •
appear bizarre at first, there is no hope” (quoted by
dhyay, 1995).
Does the CMF Mean Dualism?

The essential feature in the dualist view is the propos
physical and the mental phenomena represent two sep
tities. The extreme version of dualism was that from
He proposed that there are two kinds of substances
tans, which exhibits mental and conscious propertie
extensa, the substance of the physical world (includin
son’s body). He claimed that the two substances com
with each other by way of the pineal gland. The pine
the only brain structure that is single; all other brain
are bilateral and thus doubled. Descartes thought
pineal gland could thus account for the unitary natu
sciousness. A major distinction for Descartes was tha
tal is nondivisible and unitary, while the physical wor
ble and has extension (in other words, location in spac
But there are less extreme versions of dualism. Th
postulate separable substances. Rather, they propos
some kind of dualistic aspect in the relationship be
physical and mental worlds, which proponents claim
for the apparent nonreducibility between the two wo
is, mental subjective phenomena cannot a priori be
by any knowledge of physical events and structures;
versely, physical events (including the neuronal on
brain) cannot be described by knowledge of the acco
substances. In identity theory, the mental and the ph
nomena are regarded as two aspects of a unitary sub
is a “dual aspect” theory. There is an “inner” aspect,
tive, accessible only to the individual, and an “outer”
nal” aspect, represented in the externally observab
structures and functions of the brain. This theory see
able because there is no way of getting directly at
substrate that allegedly exhibits this double aspect. Id
ory is actually proposed as a “monist,” not a dualist
the definitions of monism versus dualism begin to los
fulness at these levels. (Remember that definitions are
designed to be helpful when considering different p
when they cease to be helpful, you should abando
least for those conditions.)
So, how does the CMF fit into this argument? T
proposed as a “property” of an emergent phenome
brain. The CMF is clearly not in the separate subs
gory of Cartesian dualism. The CMF does not exist w
brain. It emerges from the appropriate system of n
vities.
On the other hand, the emergent CMF phenomeno
lated to exhibit qualities not describable by the phys
activities that gave rise to the CMF. In a sense, that is
to conditions in the physical world, in which the pro
system are not describable by those of the compo
make up the system. (See the example I provided
benzene, the system produced by six carbon and si
atoms.) The major difference from all other syste
182 •
call this situation dualistic, you should realize that th
dualism is not Cartesian; in a sense, it also applies to a
systems.
A charge leveled at any form of dualism has been th
“ghost in the machine” (see Ryle, quoted in McGinn,
machine is the brain and the ghost is the mental cons
nomenon because it is not physically observable.
course, argued the latter is true of conscious subject
ence, whether or not it is involved as an attribute of a
a fact that subjective experience is not evident in or d
by purely physical observations of nerve cell activitie
tions between the subjective and the physical must be
by simultaneous studies of both categories.
You can, of course, ask the proponents of the “anti
gument, How do they know there is no ghost in the
The answer is that they do not really know. The eme
conscious subjective experience from nerve cell activit
mystery. If you want to refer to subjective experience
you can do so. This anti-ghost belief reminds me of
gous argument between Einstein and Stephen Hawk
Einstein disliked the proposal in quantum theory that
probabilistic rather than certain. Einstein said he did n
God played dice with the universe. Hawking’s resp
“How does Einstein know that God does not play dic
universe?”
There is another argument against mental, subje
tions that are not predictable from knowledge of th
system. Such a mind would be a loose cannon, wi
able from the physical observations of the cerebral sy
which this mind would emerge. Second, mental func
do empirically operate like a loose cannon. As a resul
cannon argument is not necessarily contradictory to
indeterminate mind.
Regardless of whether the CMF theory is valid, a
of nerve cell structures and functions can never, in its
or describe conscious subjective experience. As no
studying the brain can show you what nerve cells are
so forth, but there is nothing in that which exhibits o
any subjective experience. Also, it is possible that so
phenomena have no direct neuronal basis (see Chapt
is also possible that the conscious will does not alway
natural laws of the physical world (see Chapter 4).
We may, therefore, have to be satisfied with kno
how conscious subjective experience is related to brai
but we may not be able to explain why or how subjec
ence can be produced by brain activities any more th
explain why gravity is a property of matter. We accep
fundamental category of phenomena exists and that
to other systems may be studied without knowing w
lationships exist.
184 •
6
WHAT DOES IT ALL MEAN?
Oh God—Thy ship is so Great, and my boat is so sm
—Anonymous
The work is not yours to finish, but neither are you

to take no part in it.
—Rabbi Tarfon, in Ethics of the Fathers (quoted b

1988)
An Imaginary Dialogue between

René Descartes and the Author (BL)
René Descartes (1596–1650) is regarded as the pionee
pher who attempted to deal with the relation betw
and body in a systematic manner. An imaginary dial
him may point up how some of our present findings
have developed in comparison with the basic views of
more than 400 years ago.
Descartes started by asking, What am I absolutel
Descartes was a father of the dualist view, in which
body (including the brain) represent two separate “s
He proposed a scheme to explain how the two se
stances could interact and affect each other. These
views of the mind-body dualism have been subjecte
and vilification, especially by modern philosophers
cannot be dismissed either logically or on the basis
experimental evidence. In any case, Descartes’s reco
the distinctions between mind and the brain, and of
role of the brain for the thinking functions, provide
mental basis for later developments of how brain and
be related.
In addition, Descartes’s separation of the body (a
world generally) from any metaphysical properties an
subjective mental aspects freed physics and physiolog
a mechanistic approach. Descartes thus helped to
philosophical basis for the development of objectiv
study of the observable world.
In this imaginary dialogue, the responses by De
based, insofar as possible, on the views he expressed
writings.
BL: M. Descartes, it is an honor for me to be able t

with you some of the topics in the mind-body
ship, in which you were a pioneer. At the prese
shall try to take into account the increased fact
edge achieved by the end of the twentieth cen
ticularly our own experimental findings.
186 •
W H AT D O E S I T A L L M E A N ?
criticisms with a detailed analysis. I shall try no
with a relaxed objectivity the arguments you m
do want to note that my appearance here in yo
cannot be regarded as proof of immortality.
BL: You are, of course, well known for your insiste
starting with a basic truth of which one can ha
doubts. This was summed up in your pronounc
“Cogito ergo sum.” That is, “I think, therefore,
One could doubt whatever one thinks, but abo
thinking there can be no doubt. Or, we cannot
existence without existing while we doubt.
Now, you have been accused of emphasizing
thought as perhaps the real feature of the think
In fact, there is a recent book by an eminent ne
psychologist, Antonio Damasio (1994), with the
cartes’s Error: Emotion, Reason and the Human Bra
Damasio argues that emotional feelings are the
engine of the mind, rather than rational thinki
you clarify for us what you really meant by “th
RD: Mais oui. It is true that I emphasized rational th
a supreme judge of truth, but only in relation t
stract sciences like mathematics. However, I de
concept of “thinking” much more broadly in se
places in my writings.
I regarded thought as meaning all that we ar
ately conscious of. That includes being conscio
will, intellect, imagination, senses, emotional fe
our loves, hates, desires, joy, sadness, anger, and
W H AT D O E S I T A L L
mind. So, the distinction between my views an
Damasio is not a black-and-white one; rather, i
haps a difference in the emphasis on the role o
tions.
BL: In connection with your dictum “I think, there
am,” there is a joke that occasionally surfaces e
present. I hope you will forgive me for repeatin
now. In this joke, you go to a McDonald’s to o
hamburger. When the clerk asks you whether
mustard, you reply, “I don’t think (so)”—where
vanish from existence!
RD: Ha ha! Quel wit. Actually, I had myself raised
of possibility in my second “Meditation.” (Of
did that simply on a premise that I would ceas
in general, not specifically about mustard on a
burger.) My immediate answer to this continge
written in the Meditation, was that I would no
exist, because if I were thinking about not thin
I would have to exist to do that.
Of course, the joke depends on a semantic p
word “think.” You should adhere to my broade
of that word; to think is to be immediately cons
anything.
BL: Because you put it that way, I would like to ask
opinion of how I deal with the concept of con
perience. In my view, the essence of conscious
ence is being aware of something. The content
awareness can be anything. But being aware is
188 •
solutions to mathematical problems, can proce
sciously. That kind of thinking would then not
evidence for being consciously sure of one’s ex
RD: I think I could accept your view of conscious e
as being grounded in awareness. In a way it is c
my view of the unassailable truth—that being a
what I am really sure of and what proves I exist
ever it is that I am aware of.
BL: I like your insistence that we should reject know
grounded in authority and that we must look fo
dence to establish a truth. You further state tha
more than one way to deduce a cause for an ef
that only an experiment can decide which alter
valid. This view is remarkably close to modern
mental science, in which different hypotheses fo
ing an observation are subject to experimental
can help to indicate which hypothesis is the sup
Would you accept an application of this scienti
method to test the validity of some of your pro
RD: Well, I would have to say yes, except for axiom
ples that are untestable. The primary example o
of course, “cogito ergo sum.” I know this intui
something that is irrefutable, although we may
that to “I am aware, therefore I am.”
BL: I accept your qualification. I should add the qua
fostered by the philosopher of science, Karl Po
per insisted that it must be possible to devise an
mental test that could potentially falsify or cont
definitive truths.
RD: I like Popper’s way of making the evidence co
Well, what would you like to subject to such a
approach?
BL: Well, there is your proposal that the pineal gla
focal site in the brain where the interaction be
brain activities and the mind is arranged. Wou
please tell us how you arrived at that proposal?
RD: Yes. I realized that thought is not divisible. For
the conscious, subjective visual image is a unifi
tary “picture.” That is, the conscious image is
even though the messages from the eyes arrive
brain via many individual nerve fibers in the o
nerves. Also, virtually all the structures in the
double, one on the left side matched by one on
side. But the conscious sensory image is not do
my view of the mechanical operations of the n
system, I thought there must be a structure in
where the multitude of brain messages could b
to a single focus for interaction with the mind,
a doubled fashion. Because the pineal gland wa
structure in the brain that was not doubled, it
good candidate for the single focal site of inter
BL: The logic of your proposal is good. But it is no
possible explanation for dealing with the unita
of thought or conscious experience. Indeed, so
proposals have recently appeared. Also, we kno
pineal gland receives only a tiny fraction of the
nerve pathways. So we should have further evi
190 •
search for evidence that uniquely supported an
tory hypothesis, even though I had myself argu
that approach in principle. What would you su
BL: An obvious test would be to see what happens
neal gland became nonfunctional. In such a cas
proposal would predict the loss of virtually all m
brain interactions. Although your general dictu
ergo sum, would mean that the person could st
and be conscious, the person would lose all awa
and become nonresponsive at least to sensory m
in the brain. He would also lose the brain mess
lead the mind to perceive hunger, thirst, bodily
and so on. Furthermore, the mind/soul could n
affect brain activity, so bodily acts in response t
scious will could not take place. The person wo
come paralyzed.
I must admit that it would not be easy or eve
to deliberately destroy or inactivate the pineal g
living person. But one could at least look for po
cases of diseased pineal glands in autopsies of d
ple and relate any such diseased condition to th
ior of the individual before death.
RD: I agree that a search for cases of diseased pinea
autopsies would be desirable and would be of i
a test of the role of the pineal gland. I may not
ever, that even you modern neuroscientists hav
able to establish a mechanism for mental unity
experimental test.
BL: That is true. You proposed that the mind/soul
could exist even if there were no body substan
you noted that the existence of your thinking
one thing you have no doubt of, whereas one c
as certain about the body’s existence.
On the other hand, in a later writing, you st
the mind has its principal seat in the brain and
soul does not perceive excepting insofar it is in
And you note that the human mind is shown t
distinct from the body, and nevertheless, to be
conjoined therewith, as together to form, as it
unity. That view is surprisingly similar to mod
However, that view seems to blur the separatio
tween mind and body. Could you clarify your
on this?
RD: Mais oui. You have indeed raised a difficult poi
ever, I believe those statements of mine are no
flict with my basic position. My acceptance of
as the structure that mediates the interaction b
mind and body does not necessarily eliminate
that the two are separate entities. How the two
entities, mind and brain, accomplish the two-w
tion between them is a difficult question for w
posed an answer; I assume you will want to de
that, too.
BL: Yes, but let’s stick with the question of separat
the moment. Would you allow for the alternat
bility that mind and brain are not in fact separ
but are somehow reflections or “properties” of
192 •
know, there is now an immense amount of evid
mental, conscious processes are related to and d
on specific structures and functions of the brain
RD: Well, I suppose I cannot logically exclude the a
that you describe. However, I have noted that m
body differ in at least two fundamental ways. M
not be conceived as divisible, while body is clea
ble. Second, mind is a thinking thing while the
thing that has “extension.” That is, the location
mensions of the body can be measured. Becaus
and body do not have these fundamental attrib
common, I concluded that these two “natures”
stances” are separated one from the other. As I
the appendix to Principles of Philosophy, “two su
are said to be really distinct when each of them
without the other” (see Descartes, 1644).
BL: Well, I must respectfully argue that you cannot
know that mind and body can each exist witho
other. That uncertainty makes the view of thei
separateness a questionable one. However, I sh
push this issue further at present. I would prefe
retain your equanimity and your friendly willin
cooperate in this dialogue.
RD: Monsieur, I am, as you say, more mellow now,
promised not to become angry about opposing
will repeat, in relation to your last argument, th
tions between mind and body, the former being
ble and without extension, the latter divisible a
logically distinct and each is not describable by
However, I did recognize that brain and mind
teract; the brain is where the mind is informed
fected by perceptions and where, in turn, the m
induce the brain to control bodily actions.
In view of the immense accumulation in rec
ries of evidence that the mind is dependent on
for its manifestation, I could consider giving up
tence that the mind and body substances can e
without the other. Nevertheless, that would no
my proposal that mind and brain are separate
entities—in other words, my so-called dualism
note here that I specifically stated that I would
that this dualism indicated that the mind or so
mortal. I had no way of knowing that immort
real. That view contributed to my troubles wit
Church. However, I did point out that one can
immortality as a matter of faith.
BL: Good. I accept the logic of your argument. I w
to commend you for your prescience in sugges
is only in the brain that the mind becomes info
the sensory world. In several places you made
that even if the sensory nerve from a given are
is excited at any point along its path to the bra
mind would still perceive the sensation as loca
place of origin in the skin. That is, every sensa
ceived only by an interaction at the brain, betw
sensory message and the mind. But, subjective
194 •
terminology, I would say the sensation is subjec
ferred to a proper point in space, even though t
tern of the corresponding activities in the brain
look like the subjective image of it. I shall be di
this phenomenon at some length elsewhere in r
our experimental evidence.
RD: Well, it is gratifying indeed that my view of the
sensation is transmitted to the mind, which I de
350 years ago, still makes sense today.
BL: There is another important issue on which I wo
your comments. In your views about the mind,
pear to restrict its nature to the presence of con
experience.
RD: Mais oui. I could only be certain of the existenc
conscious thinking. As we have already discusse
thinking, I meant an immediate consciousness
thing.
BL: Yes, a clearly defensible position. However, in m
cent centuries, considerable evidence has appea
dicate that many of our mental processes are c
unconsciously, without awareness. Direct evide
appeared in the last few decades (see Chapters
But long before that, we had much convincing
and anecdotal evidence. For example, the great
mathematician Poincaré described how the solu
difficult problem suddenly appeared in his cons
mind, without his having been aware of the wa
which the solution was achieved. That is, the w
ing thing”?
RD: I must admit that the evidence for unconsciou
operations is convincing. Nevertheless, if we tu
my cogito ergo sum, it is clear that I can only
of the existence of my conscious thought. I ca
certain of the existence of a mental process of
am not aware.
On the other hand, having scientific evidenc
guish between alternative hypotheses is the be
ward truth. And so I would be willing to say th
istence of unconscious mental processes seem
probably true, based on the evidence for it, alt
cannot say that with the certainty I have about
tence of my conscious thinking.
BL: Well, on that note, let me cite some experimen
dence that directly confirms the existence of a
scious mental process. Subjects received electri
to an ascending sensory pathway in the brain.
ficiently long lasting train of electrical pulses (u
msec), they could report having a conscious se
With short trains of stimulus pulses (100 to 200
the subjects could not consciously feel any sen
But they could fairly accurately report that a st
had been delivered, even though they felt noth
an analogous kind of experiment (by Weiskran
patients who had lost conscious vision (becaus
lesion in the primary visual receiving area of t
bral cortex) could nevertheless correctly point
196 •
of unconscious mental processes, but I still thin
conclusion does not achieve the certainty I have
the existence of my conscious thinking (or feel
so on).
BL: Let me ask about your view on the existence of
RD: Oh, I think there is little doubt that we have fre
for at least some of our actions.
BL: You may be interested to hear about our experi
findings on this issue. We found that the brain
initiate and prepare for a voluntary act about 40
before the person becomes consciously aware o
her intention or wish to act. That means that c
free will does not initiate the volitional process;
initiates the process unconsciously.
RD: So is there any possibility for free will to have a
volition?
BL: Yes. Conscious intention does appear about 150
fore the motor act. That leaves sufficient time f
conscious function to intervene in the process.
provide a trigger to enable the volitional proces
completion; however, there is no direct evidenc
But there is evidence that the conscious will can
veto the process so that no act occurs. In such a
will could control the outcome. That fits with o
that we can control ourselves, something that e
tems urge us to do.
RD: I am glad to hear of that role for conscious will
how do you relate all this to the argument by t
back to your own view on what to believe. Th
can know with certainty what we ourselves are
sciously aware of. We are aware of feeling that
control the appearance of a voluntary act, incl
what to do or when to do it. That is a strong a
for actual free will. Determinism works well fo
physical world, but it is only a theory when ap
the conscious mental sphere.
RD: Again, I am delighted that this fundamental as
philosophy still has crucial merit.
BL: Well, M. Descartes, merci beaucoup for your p
tion and patience in this dialogue.
How May Our Experimental

Findings Affect How You View Yourself ?
We now know something about what nerve cells hav
ing to support or mediate the appearance of a consci
tive experience, in particular, awareness. These are u
ral activities that must be superimposed on the b
functions of a relatively normal brain. The special ne
ties center around time factors (see also Poppel, 1988
Conscious mental events appear only after a mini
tion of activations. This is as long as 0.5 sec or mo
shorter than 0.5 sec with stronger activities. Unconsci
events do not require such long-lasting activations. T
pear even with very brief neural activations, lasting as
sec or less. This describes my time-on theory, to pro
198 •
of our sensory world is delayed by a significant tim
Göteborg-Post, the leading newspaper in Göteborg, Swe
in their headline poster (when reporting on my lectu
1993), “Now it has been proven: We are all somewha
We are not consciously living in the actual present!
So, there is the unexpected evidence that there is a
delay, up to about 500 msec, in our conscious experie
sensory world. Admittedly, this was thoroughly establ
for bodily sensations, but indirect evidence makes it l
applicable to all sensory modalities.
But strangely, we are not aware of this delay. There
tive antedating of the delayed experience, back to th
the fastest response of the sensory cerebral cortex.
this “subjective referral backwards in time.” This all
feel that we are aware of a sensory signal almost im
when in fact the awareness could not have appeared
delay required by the duration of appropriate neural
developing the awareness.
On the other hand, the delay in the actual appe
awareness makes sense for other mental phenomena.
Unconscious mental events do not require such lo
neural activations. As stated above, they can appear
very brief delays lasting 100 msec or less. (Our experim
dence for my time-on theory provided direct proof for
responses to sensory signals are developed unconsciou
are evident in almost all sports activities, but also in ev
sponses to danger signals. There is experimental evi
responses in tests for reaction times (RTs) are ma
Of course, in all these cases, you can become awar
sponse or action after it has occurred. If you try
aware of the action before making it, the whole
slowed and becomes much less effective.
We may extrapolate the delay feature to all in
awareness. That would mean that all mental events w
ness would be preceded by unconscious processes th
to 500 msec before awareness. Note that there is no
process available to subjectively affect most awarene
dating can occur only for conscious sensations res
sensory signals. Therefore, all of our conscious thou
be initiated unconsciously and be delayed by up to 50
lowing their unconscious beginnings. In other words
conscious thoughts would arise unconsciously! Tha
true even for creative and complex mental operatio
that must put a fundamental qualification on our vi
all such thinking arises. It also suggests that we shou
the conditions in which free play of unconscious men
may proceed.
There could not be a continuous stream of consc
there were a substantial delay in the production of a
event. Conscious events would have to appear disco
The fact that we ordinarily experience being consciou
tinuous manner can be explained by an overlap in t
ance of multiple conscious events.
The modification or distortion of a subjective exp
well-accepted phenomenon in psychiatry and psycho
proposed that emotionally charged sensory images an
200 •
To modify a sensory image, there would have to be
able after the sensory messages arrived at the cereb
During such a delay, the brain could react to the imag
erate neural patterns to modify it before the subje
consciously aware of the image. The physiological re
of a delay for development of awareness provides an
nity for unconscious cerebral processes to modify the
a subjective experience. Indeed, we showed that stim
sensory cortex, applied hundreds of msec after a stim
skin, could retroactively either depress or enhance th
experience of that skin stimulus.
Finally, there is the discovery that the brain unconsc
tiates the volitional process well before a person beco
of an intention or wish to act voluntarily. That result
a profound impact on how we view the nature of fre
on issues of personal responsibility and guilt (see Cha
The various examples described here show how th
tor for awareness can have a profound impact on our
mental life.
At this point, the comments of philosopher David
thal (2002) are appropriate. These appeared in the
issue of Consciousness and Cognition, a special issue d
timing relations between brain and world that was pr
voted to commentaries on my work in this area. (T
number of other critical and supportive commentar
work in this issue. I have now had the opportunity to
publish my replies to most of them; see Libet, 2003)
Rosenthal states that he has “little to say in this
which that conflict seems to arise.” The “picture” Ro
in mind refers to our experimental evidence for delay
ing awareness, in the case of conscious sensation,
appearance of the conscious wish or urge to act
the prior unconscious initiation of the volitional pro
brain).
Rosenthal’s chief argument is that the general com
picture of such events is in conflict with our experim
ings. He believes that makes our experimental conc
likely to be valid. He does note that a person’s ability
performance of a volitional urge to act “might mitiga
flict somewhat, but only if we have independent ev
this conscious veto does not itself occur after a no
neural veto.” I have analyzed this latter issue at so
(Libet, 1999), and showed that the evidence is comp
the view that the conscious veto arises without the p
nonconscious processes having made the final decisio
Rosenthal goes on to argue that “our commonse
plainly accommodates nonconscious volitions.” But
ment seems to defeat his point regarding the commo
adox, which is that our findings “compromise our se
agency.” He then claims that our findings of unconsc
tion of the volitional process are actually quite comp
our commonsense picture! If you want to accept u
initiations of voluntary acts as not in conflict with yo
of free agency, you can certainly do so. But then you
trouble with the concepts of responsibility and guilt f
ations over which you have no conscious control.
202 •
that many philosophers share: Untested philosophic
and speculative views may be proposed as if they we
of contradicting conclusions that are based on experim
dence. As scientists, we cannot accept that. Proposed
theories are valuable only if they help explain the
when they contradict the data. A “commonsense” vi
supercede hard-gained experimental data. Experimen
eries often present counterintuitive results and infer
are original and creative. There is perhaps nothing m
terintuitive and in conflict with commonsense than
theory. Yet it is regarded as a major pillar of physic
rectly predicts experimental observations.
How Should We View Self and Soul?

We come finally to consider the nature of the self an
in relation to conscious functions of the brain. Are th
cases of conscious processes, or are they in separate
that are independent of conscious processes generally
emerge from appropriate activities of nerve cells in
Or are they entities that are separate from the physica
a Cartesian sense? On this last point, we must reco
there is no evidence to support the concept of sepa
status, which can only be a metaphysical belief. I ther
further discussion on the view that the self and th
emergent phenomena of brain activity.
Certainly, these phenomena can be altered or ab
pharmacological agents and by pathological chang
There have been many views expressed regarding
origins, and meanings of self (von Weiszaker noted
characteristic of psychology not to really ask wha
quoted in Del Guidice, 1993). Most of these represent
ical analyses and interesting speculations. I would li
the discussion here to the simplest phenomenal
self—in other words, to those features that are report
ences by most human beings. A primary status for s
able phenomena is adopted as the basis of the book b
per and John C. Eccles (1977), The Self and Its Brain, a
a reversal of the more common view of a brain and
The simplest phenomenological view of self is the
feeling we each have of being our own person, wit
personal identity. Neurologist Antonio Damasio (199
a distinction between a transient self (that is continuo
ated as a result of one’s ongoing experiences of t
world), and an autobiographical self (based on mem
periences and so on). Damasio refers to the forme
self as a “core self.” I prefer to reserve the term “core
enduring aspect of self, the personal identity one
when there are significant and even extreme changes
tent of one’s capacities for conscious experience.
The resistance to change, for what I am calling the
quite remarkable. Even after temporary loss of con
due to a variety of causes, people know who they a
turning to consciousness. The same phenomenon oc
person awakes from ordinary sleep, general anesthesi
long coma. Lesions of the cerebral cortex can be quit
204 •
between the two cerebral hemispheres have been
hemisphere can possess knowledge of events not avail
other hemisphere. Yet these people show no eviden
ficulties with their personal identities. They do not co
having multiple personalities; they still feel they are th
dividual they were before the splitting of the intercon
The same is true even in patients in whom a who
hemisphere has been excised or destroyed by patholog
tions such as a large tumor. Their awareness of person
is continuous in spite of drastic losses in mental func
as paralysis and loss of sensation on one side or a s
if the dominant (usually the left) hemisphere is go
viewed a video of such a patient who seemed co
aware of, and responsive to, an interviewer. Yet he kn
well what his deficiencies were.
Further, patients with bilateral lesions in the hip
structures in their temporal lobes lose the ability to
explicit memories, but they retain memories of even
curred before the injury. These patients also appea
who they are and are even aware of their loss of me
mation.
Even complete amnesia concerning past history or
name does not appear to destroy the sense of self. O
there is no autobiographical self during the amnesia,
reappear when the patient recovers from the amnes
cently reported case, a young woman suffered comple
after being struck into a coma by an automobile, but
cover full consciousness. After two years in amnesia, t
recall her Social Security number, she was identified.
a rapid restoration of memories concerning her pas
ing of personal identity was not permanently destro
loss of her autobiographical self for two years!
How Does Self Relate to Awareness?

I proposed in an earlier chapter that awareness is a fu
phenomenon, with its own unique requirements of n
ities. I also proposed that the different contents in
may account for the variety of conscious experienc
our having to regard the different classes of experienc
ent kinds of awareness. I include the phenomenon of
view. That is, the experience of selfhood may represen
content added to awareness. Theorists have produce
of selves to account for the actual variety in phenom
displays of a self. It is simpler to view these varietie
variations in the contents of basic awareness rather th
ent levels and kinds of awarenesses.
There is a puzzling question in this view of the
awareness. When there is awareness of some mental
whether these contents include the feeling of self or
sory stimuli, who is aware of that mental content? Th
there is a personal entity that experiences these con
popular with most philosophers and neuroscientists
personal entity cannot be viewed as a specific neural
tion in some localized part of the brain. Large lesions
of the cerebral hemispheres do not abolish awarene
sonal identity, though it is true that a lesion in the m
206 •
ness. But these structures are best regarded as nec
keeping the cerebral cortex awake, rather than as c
ficient for the whole experience. We are forced to
more global representation of consciousness and the
produced by whatever brain areas remain sufficient
lesions.
I come back to my proposed conscious mental field
which almost any part of the cortex may contribute, a
tial answer to our question: A unified experience of
is the postulated attribute of the CMF. See Chapter
scription of CMF and an experimental design to test t
The CMF would provide the kind of global attribut
for subjective unity and for the “who” that is aware
contents.
Are Unconscious Mental Processes Part of Self ?

Much of our mental life proceeds unconsciously, and
mental processes can affect succeeding conscious
Some of the simplest examples of this principle lie i
ings of Shevrin (1973), using very brief visual stimu
msec) whose contents were not consciously discern
subjects. Shevrin and Dickman (1980) showed that th
of these unconscious inputs distinctly affected the su
lection of word responses from a list of new words. T
selected new words showed an association with the
served” unconsciously. So these unconscious percep
an impact on later mental operations. Analogous re
found with patients after recovery from general
sponses after recovery.
Unconscious mental processes are in fact unique
individual. For example, a mathematician may solve
unconsciously, something that someone else does not
seem appropriate, then, to regard one’s unconscious
as belonging to, and characteristic of, one’s self. Ho
does not have direct experiential access to the uncons
tal processes, although these processes may have an
how we view ourselves consciously. Chapter 4 con
question of responsibility and guilt for unconsciou
processes. In my view, the crucial point is that we ha
control over the actual performance of our unconsc
ated volitional process. Hence, we are responsible fo
scious control choices, not for our unconsciously init
that precede our conscious decisions.
Is the Feeling of Self a Singular Experience?

The most direct challenge to the singular uniquenes
sonal self comes from the alleged existence of multi
ality disorder (MPD). In MPD, a patient appears to ex
than one personality at different times. There seems
personality that prevails most of the time, and it rem
sole personality when therapy successfully eliminates
The various personalities can be strikingly different
behaviors virtually opposite to those of the main
Some readers may remember the old movie The Th
Eve, adapted from a nonfiction book of that title
Cleckley. There was, of course, Stevenson’s Dr. Jek
208 •
Sybil based on clinical reports by Sybil’s psychiatris
Wilbur. The case appeared to be a documented e
MPD, and thus helped to establish a diagnosis of MPD
uine disorder in psychiatric practice.
However, the validity of the Sybil story and diagno
riously questioned by Herbert Spiegel (1997), a psych
expert in hypnosis. Spiegel had considerable direct co
Sybil for several years, as an adjunct to the therapy by
a lengthy published interview (The New York Review
April 24, 1997), Spiegel called Sybil’s diagnosis of
question. He discovered that Sybil was highly hyp
During an age-regression under hypnosis, Sybil rep
usual events of a child, but not the stories of parental
Wilbur got. In one such session, Sybil asked Spiegel if
her to be Helen, a name that Wilbur had given her fo
she had. Indeed, Sybil felt an obligation to becom
personality; but with Spiegel, she preferred not to “b
There were many further indications to support Spe
that the MPD phenomenon of Sybil was an artifact th
ated by Wilbur. Spiegel realized that Wilbur was not s
knowledgeable about hypnotic manifestations and th
had been reifying a memory of some kind and conver
a personality. Spiegel suggested that any highly hypno
tient who has been diagnosed as MPD is almost cert
real case of MPD. Other MPDs who are not highly hy
can also be misdiagnosed for other reasons.
On the other hand, Spiegel does note there are “p
transient disassociations (in their views of self ), where
We see, then, that the challenge to the single-self
the reports of MPD is, at best, a doubtful one. Howe
rary loss of personal identity can apparently occur. B
covery from that loss the person again feels she is the
son as before.
Does a Split Brain Affect Personhood?

In a split-brain patient, the commissures, which are t
of nerve fibers that connect the two cerebral hemisp
been surgically cut through to relieve certain types
“[M]ost of the conscious experience generated within
sphere becomes inaccessible to the conscious aware
other” (Sperry, 1985). A series of studies by Sperry e
showed that “these people live with two largely se
and right domains of inner consciousness. Each hem
be shown to experience its own private sensation
thoughts, and memories, which are inaccessible to aw
the other hemisphere.” All the tests indicated tha
hemisphere is conscious and, except for lacking an
speak, it exhibits thoughtful, logical, and emotiona
separate from but equal in quality to those of the
sphere (see also Bogen, 1986; Doty, 1999). (There hav
agreements with this view; see Gazzaniga, 1998.)
All this evidence indicates that it is possible for the
to be two in nature. Yet the split-brain patients appea
gle-minded, unified individuals. They do not comp
loss of personal identity or of split experiences. Sp
offers some explanations for this situation. Awaren
210 •
commissurotomies. Also, there are bilateral represen
the face, for audition, and for other systems that are n
dent on the upper commissures. Division of the visua
be made ineffective by having both eyes roaming over
visual field.
So it would seem that, although some aspects o
scious self can be split in these patients, the consci
largely unified. That unified self is especially predo
normal people.
Do Identical Twins Have One or Two Separate S

Identical twins stem from a single fertilized egg and h
cal genetic makeup in all their cells. However, the ex
genes is dependent on the conditions of brain develop
the effects of environment through life. Identity of ge
fore, does not necessarily mean identity of the actua
als. Nevertheless, identical twins who have been sep
reared in different environments exhibit remarkable
in mental outlook, choices of mates, autos, behavio
as their physical appearances. But identical twins fee
unique persons. Each knows who she is and does n
herself with the other twin. Identical twins do not,
provide an example of a split self.
When, then does a unique self appear? Some rel
tems hold that a self or soul is “attached” to the fertil
conception. But identical twins start out as a single
egg, yet they develop into two independent selves.
seem more likely that some form of a self appears w
Some computer enthusiasts, especially those work
ficial intelligence, have expressed a belief that comput
conscious. They feel that if and when a computer is
as the human brain, and can behave in a way not dist
from that of a person, the computer should be regard
tionally equivalent to a human person. In the fam
paradigm, this would happen when a computer behi
could give responses indistinguishable from those o
behind the screen (of course, there is no guarante
would happen).
There have been a number of physical and philosop
ments made in opposition to such a view. The ma
physicist Roger Penrose (1989) has argued that compu
ways operating with a programmed algorithm. But
conscious mental functions can be nonalgorithm
thus fundamentally different from computer functio
“demonstrates that neither quantum theory nor class
shed any light on the question of determinism vs. fre
states that “our present science is incomplete” (see
Palm, 1991). Philosopher John Searle (1980) points
computer can respond to questions based on its pro
for the syntax of the language. It can thus appear to r
a human person. But, unlike the person, the comput
understand the meaning of the language. The comp
programmed for syntax, but not for the semantics
guage. This distinction between syntax and semantic
portant issue for consciousness, was expressed earlier
the philosopher Karl Popper.
212 •
the computer is therefore also conscious, like a pe
proof of this statement lies in a simple logical argume
dealing here with two different systems: A (the compu
(the person). A and B are known to be different in ma
for example, in the materials of which they are constr
two different systems, A and B, are found to give id
havioral responses to questions (if the Turing para
does work). Does that mean that the two systems are a
cal in other characteristics, like the possession of cons
The answer to that is no, based on simple rules of log
if system A exhibits X and system B exhibits X, it does
that both systems will exhibit Y (even if one of them
hibit Y). Systems that are similar in one respect need n
lar in other properties.
Such a logical fallacy also applies to another relat
tion. It has been suggested that we might replace each
in the brain with a silicon chip that performs all of
functions. If we could do this for the whole brain,
have an instrument that could perform functions ind
able from those of the original brain. This zombie, ac
some, would also be conscious. But, again, this is a di
tem from the original brain and it cannot be said to sh
properties of the normal brain. The brain is struct
functionally different from a system of silicon chips.
What Is Meant by the Soul?

It is not easy to distinguish the meaning of soul fro
self. For many people, soul appears to refer to a mor
meaning and feeling than does self. There is also a te
present, a metaphysical belief: It cannot be falsified
dence.
Just as the phenomenon of self may be a special cas
of content in basic awareness, so may a phenomen
be such a case, if indeed there are real differences be
and soul.
Neither self nor soul need be physical substance
they emerge from physical activities of nerve cells
nomenological feeling of their existence may be ba
cial kinds of neural processes. Feelings of self and s
destroyed by extensive neural damage, as in advan
mer’s disease or in Creutzfeldt-Jacob disease. This
even when awareness has not been completely elim
feelings of self and soul do require a sufficiently stru
functional brain.
Soulful feelings have strong meanings for many
should be taken to be serious phenomena based on a
neuronal activities. Soulful feelings are especially im
and expressed in, music, art, literature, and, of cours
and spiritual activities. Such phenomenal feelings sho
dismissed lightly, without convincing counterevidenc
As an example of this view of the soul, I quote fro
ings of the eminent novelist Saul Bellow. Bellow can
cused of basing his views on a religious indoctrinatio
very secular person. The following quotes are from
by Leon Wieseltier (1987). Writing of his novel Her
(1987) notes that “a Ph.D. from a good American
[Herzog] falls apart when his wife leaves him for an
What is he to do in this moment of crisis, pull A
214 •
the soul. It may be difficult to find because by midlife
grown, and some of the wildest thickets surroundin
out of what we describe as our education. But the ch
ways there, and it is our business to keep it open, to h
to the deepest part of ourselves—to that part of us wh
scious of a higher consciousness, by means of which
final judgments and put everything together. The ind
of this consciousness, which has the strength to be i
the noise of history and the distractions of our imm
roundings, is what the life struggle is all about. The s
find and hold its ground against hostile forces, some
bodied in ideas which frequently deny its very exis
which indeed often seem to be trying to annul it a
Wieseltier comments, “These wise and beautiful sent
ture the objective of Bellow’s book perfectly.” And “B
discerned the far-reaching difference between intell
and contemplative life. And also that there are sign
ments of a modern intellectual’s education that mu
learned if a beginning is to be made.” Obviously, Bell
are opposed to those of many theorists, especially
determinists.
There have been many other expressions against th
ist determinist doctrine that any feelings of self, sou
will are illusions; that we are automatons governed by
rable adherence to established physical laws; and tha
edge of the structures and functions of the elements in
will tell us everything about conscious experience an
festations. The latter view is known as reductionism.
The issues of determinism, reductionism, and free
In addition to my arguments in Chapters 4 and 5, ther
of Sperry (1985), Popper and Eccles (1977), and of m
cists.
The University of Cambridge physicist Sir Bria
(1992) notes that if a theory of everything “meant onl
rial world it would not be so bad, but it also includes
mind.” But “the very ground rules of science,—preclu
ing an explanation for my consciousness, the one ph
of which I am absolutely certain.” I would qualify th
ring to our own experimental research on conscious
in these scientifically conducted studies, reports of co
perience were accepted as valid information about
subjective experiences (see Chapters 1 and 2).
The physicist Brian D. Josephson (1993) wrote a r
symposium dealing with reductionism in science a
He reported that physicists, mathematicians, and p
in that symposium considered reductionism to be un
and misleading.
In any case, the phenomena of self or soul, and
tially causative role they may have in mind-brain in
are not made meaningless by any determinist, ma
reductionist theory.
Is There Life after Death?

Some semblance of evidence for life after death has
the reports of persons who have had near-death e
People who experience cardiac arrest go through sta
of cerebral functions. There is an initial phase in wh
216 •
brain is resumed during that time. That is, the damage
curring during this time is potentially reversible. With
riods of cardiac arrest, damage becomes progressive
ible.
Cardiac arrest may appear suddenly, as when the ma
ing heart muscle (the ventricles) suddenly becomes dis
(fibrillation) and ineffective as a pump of the blood.
page of circulation of blood to the brain results in a p
loss of function for different parts of the brain. Th
(and cerebellar) cortex ceases activity within 5 to 10 s
consciousness and then the electrical rhythms (brain
electroencephalogram) disappear in that time. The s
parts of the brain are a bit more resistant; but after ab
even the lower brain stem goes out, with a loss of bre
other bodily controls that have their centers in the me
spinal cord, and the simple reflexes it mediates, sur
longer (for 1 to 2 min).
When all these brain activities are lost, the person
all intents and purposes, to be dead. However, if the
be induced to resume effective pumping within a max
riod of about 5 min, the various functions of the br
cover, though it may take weeks for full recovery. Du
min period of arrest, metabolic degradation of nerve
on because of the absence of oxygen and glucose, w
vides cellular energy, and the accumulation of met
and around neurons. The other organs of the body,
the heart, can withstand much longer periods of circ
rest before the effects become irreversible. It is thus
brain. When the brain is irreversibly dead, an ind
never recover from a coma.
Persons who have recovered from cardiac arrest ha
having had experiences during the time the brain wa
ally inactive or “dead.” A common report is of floatin
the operating room and watching the doctors and
tempting resuscitation. Another one is of the patient
he is proceeding into a tunnel with a bright light
These reports may include a feeling of peacefulness
with the tunnel and light.
What can we make of such reports? If the brain
pletely inactive (“dead”) when these reported expe
curred, we might have some impressively convincin
of extracorporeal psychic existence. But there are so
ties with these data. First, it is astonishing that a pe
remember such subjective experiences for later repo
sumably, the brain structures that mediate formation
ory were also inoperative during the cardiac arrest. S
sons who report floating about and observing the re
process could be reporting recollections of scenes, o
imagined, that were formed before the cardiac arrest
may question how thoroughly the condition of co
sence of brain functions was established in persons
reported experiences. For example, in some early wo
with my colleagues) established that a clap of hand
an initial “primary evoked electrical response” at th
cortex. This response appeared for a time after the sp
electrical rhythms associated with consciousness d
218 •
swer to this profoundly important issue of an afterlife
ficult to achieve. If a subject with a cardiac pacemake
available for a special experiment with cardiac arrest fo
a rigorous test situation could be arranged. (This co
done using a patient in the emergency room with sp
cardiac arrest.) The test would involve turning off
maker to produce a cardiac arrest lasting 3 to 4 min, w
the limits for recovery of brain function. Before the p
brought into the room, unusual and strange images a
would be introduced but covered up in the room. Th
be exposed after the loss of brain functions in the card
When active resuscitation to restore cardiac beating a
ing of blood was begun, the medical personnel woul
ately remove the special gowns and images and the v
wall pictures would be reapplied.
When the patient was again conscious and respo
could be asked to report any experiences that happen
the procedure. If the patient reported having floated
room and observing the medical personnel, she coul
to describe the backs of those personnel and anything
about the room’s walls and so forth. If the patient co
ported seeing the secreted indicators, that would be r
convincing for the validity of the report. However, if
reported seeing the usual environment of the room
the secreted one, the report’s validity would colla
other explanation of what is reported in the “after de
tion would have to be sought.
Of course, the difficulty with this experiment is tha
Another workable design for testing whether th
communicate with the living has been activated by
tists at the University of Arizona (Schwartz and Russ
work was described by Ann Japenga (1999). This de
normal individuals to participate. Each participant st
sage known only to him in a computer. The compute
the message into code. After death, the participant
cates by telepathy with an arranged person who is stil
phrase that unlocks the code is presumably given b
person to the living colleague.
Actually, this kind of experiment has, I believe,
ducted and failed. In the early 1900s, the great magici
arranged for a similar kind of test of whether he wo
to communicate with a living person after his own d
dini died not long thereafter, but his widow and frien
that they did not receive any messages from him.
There have of course been claims by some who
selves mediums that they can call up a speaking ghos
individual, but these demonstrations in séances hav
been exposed as fraudulent when investigated by an e
I have no objection to the possibility of some kin
ingful existence of a soul after death. In fact, such
could make the prospect of death profoundly more
However, we still have not gone beyond metaphysica
this issue. As we have seen, the question is extremely
answer in any scientifically convincing manner.
For the present, then, we may accept the concept
as based on phenomenologically meaningful experien
experiences do not prove there is an actual entity, th
220 •
Ryle’s attack is based on his belief that we are just
How does Ryle know there is no ghost in our cerebra
The fact is he does not know. There is no direct evi
contradicts the possible existence of a Cartesian-type
there is also no evidence that contradicts a nonphysica
enon that is not separable from the brain (as in my
ory). Nor is there evidence that confirms it, as yet.
In Conclusion
Let me repeat what I said at the beginning: Our subje
life is what really matters to us as human beings. Yet
and understand relatively little of how it arises and ho
tions in our conscious will to act. We do know th
only life we know, the physical brain is essential to
mately involved in the manifestation of our conscious
experience.
In this book, I have introduced some of the exp
progress toward discovering the significant physica
activities that mediate nonphysical conscious subject
ences. I emphasized our research, partly to give you
into how such discoveries are made and also to prov
for the conclusions and inferences I made from the
Also, our research provides rare findings into the relati
tween neural activities and subjective experience, find
on direct intracranial studies of neural stimulations a
ings. It has allowed us to establish the causal nature o
ships, not merely the correlations.
We discovered that a time factor appears to be a sig
would appear to have a profound effect on how w
mental selves. If all conscious awarenesses are prece
conscious processes, we are forced to conclude that
actually live in the present and that unconscious proc
predominant role in the production of our conscio
found that this can be extended even to an unconsc
tion of a voluntary act and appears to restrict the r
will to controlling the performance of actions. We
seen that subjective experiences of all kinds involv
tive referral of the responsible brain activities into
thoughts that give a conscious order and meaning to
cated neural activities that elicit them.
Our surprising discoveries could not have been ma
the intracranial access we had to certain parts of the b
laborative human subjects.
My chief concluding hopes are: (1) that you hav
our experimental discoveries on the mind-brain pr
that you recognize how those discoveries may affect
about your own mental experiences; and finally (3) t
fied experimental neurosurgery group will carry ou
priate test of my unified cerebral mental field theory
ter 5). I have already provided an experimental design
test. The result of the proposed test could falsify an
the theory. But a positive result, one showing subjecti
nication and intentional actions without any neural co
would have a profound impact on how we view the
conscious experience and on neuroscience generally.
222 •
BIBLIOG
BIBLIOGRAPHY
Amassian, V. E., M. Somasunderinn, J. C. Rothswell, J. B. Crocco,

Macabee, and B. L. Day. 1991. Parasthesias are elicited by si
magnetic coil stimulation of motorcortex in susceptible hu
114:2505–2520.
Baars, B. J. 1988. A Cognitive Theory of Consciousness. Cambridge, E

Cambridge University Press.
Barbur, J. L., J. D. G. Watson, R. S. J. Frackowiak, and S. Zeki. 199

visual perception without VI. Brain 116:1293–1302.
Bellow, S. 1987. More Die of Heartbreak. New York: Morrow.
Bennett, H. L., H. S. Davis, and J. A. Giannini. 1985. Nonverbal res

intraoperative conversation. British Journal of Anaesthesia 57
Berger, H. 1929. Über das electrokephalogram des menschen, Arch

Psychiatrie u. Nervenkrankheit 87: 527–570.
Berns, G. S., J. D. Cohen, and M. A. Mintun. 1997. Brain regions r

novelty in the absence of awareness. Science 276: 1272–1275.
sciousness & Cognition 4(1):52–62.
Bonke, B., P. I. M. Schmitz, F. Verhage, and A. Zwaveling. 1986. A

study of so-called unconscious perception during general
British Journal of Anaesthesia 58:957–964.
Bower, B. 1999. Science News 156:280.
Buchner, H., M. Fuchs, H. A. Wischmann, O. Dossel, I. Ludwig,

and P. Berg. 1994. Source analysis of median nerve and fin
lated somatosensory evoked potentials. Brain Topography 6
Buchner, H., R. Gobbelé, M. Wagner, M. Fuchs, T. D. Waberski,

mann. 1997. Fast visual evoked potential input into human
Neuroreport 8(11):2419–2422.
Burns, B. D. 1951. Some properties of isolated cerebral cortex in t

unanesthetized cat. Journal of Physiology (London) 112:156–17
———. 1954. The production of after-bursts in isolated unanesth

bral cortex. Journal of Physiology (London) 125:427–446.
Burns, J. 1991. Does consciousness perform a function independe

brain? Frontier Perspectives 2(1):19–34.
Buser, P. 1998. Cerveau de soi, cerveau de l’autre [One’s brain and th

another]. Paris: Odile Jacob, see pp. 30–73.
Chalmers, D. J. 1995. Facing up to the problem of consciousness.

Consciousness Studies 2(3):200–219.
———. 1996. The Conscious Mind. New York: Oxford University P
Churchland, P. M., and P. S. Churchland. 1998. On the Contrary: C

1987–1997. Cambridge, Mass.: MIT Press.
226 •
BIBLIOGRAPHY
tional Neurology. Los Altos, Calif.: Lange Medical Publishers
Clark, R. E., and L. R. Squire, 1998. Classical conditioning and bra

the role of awareness. Science 280:77–81.
Cooper, D. A. 1997. God Is a Verb: Kabbalah and the Practice of Mysti

New York: Penguin Putnam.
Crawford, B. H. 1947. Visual adaptation in relation to brief condit

uli. Proceedings of the Royal Society Series B (London) 134:283–
Crick, F. 1994. The Astonishing Hypothesis. London: Simon and Schu
Crick, F., and C. Koch. 1998. Consciousness and neuroscience. Cer

8(2):92–107.
Cushing, H. 1909. A note upon the faradic stimulation of the post

gyrus in conscious patients. Brain 32:44–53.
Damasio, A. R. 1994. Descartes’ Error. New York: Penguin Putnam.
———. 1997. Neuropsychology. Towards a neuropathology of em

mood. Nature 386(6627):769–770.
———. 1999. The Feeling of What Happens: Body and Emotions in th

Consciousness. New York: Harcourt Brace.
Del Guidice, E. 1993. Coherence in condensed and living matter. F

spectives 3(2):6–20.
Dember, W. N., and D. G. Purcell. 1967. Recovery of masked visu

inhibition of the masking stimulus. Science 157:1335–1336.
Dennett, D. C. 1984. Elbow Room: The Varieties of Free Will Worth W

Cambridge, Mass.: Bradford Books (MIT Press).
BIBLIOG
ness. Ciba Foundation Symposium #174. Chichester, Engla
Wiley and Sons.
Dennett, D. C., and M. Kinsbourne. 1992. Time and the observer

and when of consciousness in the brain. Behavioral and Bra
15:183–247.
Descartes, R. 1644 [1972]. Treatise of Man, trans. T. S. Hall. Cambr

Harvard University Press.
Doty, R. W. 1969. Electrical stimulation of the brain in behaviora

nual Reviews of Physiology 20:289–320.
———. 1984. Some thoughts and some experiments on memory.

Neuropsychology of Memory, eds. L. R. Squire and N. Butter
Guilford.
———. 1998. Five mysteries of the mind, and their consequences

Neuropsychologia 36:1069–1076.
———. 1999. Two brains, one person. Brain Research Bulletin 50:4
Drachman, D. A., and J. Arbit. 1966. Memory and the hippocamp

is memory a multiple process? Archives of Neurology 15(1):52
Eccles, J. C. 1966. Brain and Conscious Experience. New York: Sprin
———. 1990. A unitary hypothesis of mind-brain interaction in c

tex. Proceedings of the Royal Society B (London) 240:433–451.
Echlin, F. A., V. Arnett, and J. Zoll. 1952. Paroxysmal high voltage

from isolated and partially isolated human and animal cer
Electroencephalography & Clinical Neurophysiology 4:147–164.
228 •
BIBLIOGRAPHY
stereotoxic technique in man allowing multiple spatial and
approaches to intracranial targets. Journal of Neurosurgery 11
Feynman, R. 1990. In: No Ordinary Genius, ed. C. Sykes. New York

Norton, p. 252.
Franco, R. 1989. Intuitive science. Nature 338:536.
Freud, S. 1915 [1955]. The Unconscious. London: Hogarth Press.
Gazzaniga, M. S. 1998. Brain and conscious experience. Advances i

77:181–192, plus discussion on pp. 192–193.
Goff, G. A., Y. Matsumiya, T. Allison, and W. R. Goff. 1977. The sc

phy of human somatosensory and auditory evoked potenti
encephalography & Clinical Neurophysiology 42:57–76.
Goldberg, G., and K. K. Bloom. 1990. The alien hand sign: localiza
lateralization and recovery. American Journal of Physical Med
habilitation 69:228–230.
Goldring, S., J. L. O’Leary, T. G. Holmes, and M. J. Jerva. 1961. Dir

of isolated cerebral cortex of cat. Journal of Neurophysiology
Gray, C. M., and W. Singer. 1989. Stimulus-specific neuronal oscill

entation columns of cat visual cortex. Proceedings of the Na
emy of Sciences, U.S.A. 86:1698–1702.
Green, D. M. and J. A. Swets. 1966. Signal Detection Theory and Psy

New York: John Wiley and Sons.
Grossman, R. G. 1980. Are current concepts and methods in neur

equate for studying the neural basis of consciousness and m
BIBLIOG
conscious awareness. Experimental Brain Research 126:128–13
Haggard, P., and B. Libet. 2001. Conscious intention and brain ac

of Consciousness Studies 8:47–64.
Halliday, A. M., and R. Mingay. 1961. Retroactive raising of a sens

old by a contralateral stimulus. Quarterly Journal of Experim
ogy 13:1–11.
Hawking, S. 1988. A Brief History of Time. New York: Bantam Boo
Hook, S., ed. 1960. Dimensions of Mind. Washington Square: New

sity Press.
Hubel, D. H, and T. N. Wiesel. 1962. Receptive fields, binocular i

and functional architecture in the cat’s visual cortex. Journ
ogy (London) 160:106–134.
Ingvar, D. H. 1955. Extraneuronal influences upon the electrical a

lated cortex following stimulation of the reticular activatin
Acta Physiologica Scand 33:169–193.
———. 1979. Hyperfrontal distribution of the cerebral grey matt

in resting wakefulness: on the functional anatomy of the
state. Acta Neurologica Scand. 60:12–25.
———. 1999. On volition: a neuro-physiologically oriented essay.

tional Brain: Towards a Neuroscience of Free Will, eds. B. Lib
man, and K. Sutherland. Thorverton: Imprint Academic, p
Ingvar, D., and L. Phillipson. 1977. Distribution of cerebral blood

dominant hemisphere during motor ideation and motor p
Annals of Neurology 2:230–237.
230 •
BIBLIOGRAPHY
Jasper, H., and G. Bertrand. 1966. Recording with micro-electrode
stereotaxic surgery for Parkinson’s disease. Journal of Neuro
24:219–224.
Jeannerod, M. 1997. The Cognitive Neuroscience of Action. Oxford: B
Jensen, A. R. 1979. “g”: outmoded theory of unconquered frontier

Science and Technology 2:16–29.
Josephson, B. D. 1993. Report on a symposium on reductionism in

culture. Frontier Perspectives 3(2):29–32.
Jung, R., A. Hufschmidt, and W. Moschallski. 1982. Slow brain po

writing: The correlation between writing hand and speech
in right-handed humans. Archiv fur Pschiatrie und Nervenkran
232:305–324.
Kaufmann, W. 1961. Faith of a Heretic. New York: Doubleday.
Keller, I., and H. Heckhausen. 1990. Readiness potentials precedin

ous acts: voluntary vs. involuntary control. Electroencephalo
Clinical Neurophysiology 76:351–361.
Kihlstrom, J. F. 1984. Conscious, subconscious, unconscious: a cog

spective. In: The Unconscious Reconsidered, eds. K. S. Bowers
Meichenbaum. New York: John Wiley and Sons.
———. 1993. The psychological unconscious and the self. In: Expe
Theoretical Studies of Consciousness. Ciba Foundation Sympo
Chichester, England: John Wiley and Sons.
———. 1996. Perception without awareness of what is perceived,

without awareness of what is learned. In: The Science of Co
BIBLIOG
Kornbuber, H. H., and L. Deecke. 1965. Hirnpotential ändrungen
Willkürbewegungen und passiven Bewegungen des Mensc
Bereitschaftpotential und reafferente potentiale. Pflügers Ar
Kristiansen, K., and G. Courtois. 1949. Rhythmic electrical activit

lated cerebral cortex. Electroencephalography and Clinical Ne
1:265–272.
Laplace, P. S. 1914 [1951]. A Philosophical Essay on Probabilities, tran

Truscott and F. I. Emory. New York: Dover.
Lassen, N. A., and D. H. Ingvar. 1961. The blood flow of the cere
determined by radioactive Krypton 85. Experientia 17:42–43
Libet, B. 1965. Cortical activation in conscious and unconscious e

Perspectives in Biology and Medicine 9:77–86.
———. 1966. Brain stimulation and the threshold of conscious ex

Brain and Conscious Experience, ed. J. C. Eccles. New York:
Verlag, pp. 165–181.
———. 1973. Electrical stimulation of cortex in human subjects a

sensory aspects, In Handbook of Sensory Physiology, ed. A. Ig
Springer-Verlag, pp. 743–790.
———. 1980. Commentary on J. R. Searle’s “Mind, Brains and Pr

havioral and Brain Sciences 3:434.
———. 1985. Unconscious cerebral initiative and the role of cons

voluntary action. Behavioral and Brain Sciences 8:529–566.
———. 1987. Consciousness: conscious, subjective experience. In

of Neuroscience, ed. G. Adelman. Boston: Birkhäuser, pp. 27
232 •
BIBLIOGRAPHY
———. 1993a. Neurophysiology of Consciousness: Selected Papers and
by Benjamin Libet. Boston: Birkhäuser.
———. 1993b. The neural time factor in conscious and unconscio

Experimental and Theoretical Studies of Consciousness. Ciba Fo
Symposium #174. Chichester, England: John Wiley and Son
146.
———. 1994. A testable field theory of mind-brain interaction. Jou

sciousness Studies 1(1):119–126.
———. 1996. Solutions to the hard problem of consciousness. Jou

sciousness Studies 3:33–35.
———. 1997. Conscious mind as a force field: a reply to Lindhal &

Journal of Theoretical Biology 185:137–138.
———. 1999. Do we have free will? Journal of Consciousness Studies 6
———. 2001. “Consciousness, free action and the brain”: commen

John Searle’s article. Journal of Consciousness Studies 8(8):59–
———. 2003. Timing of conscious experience: reply to the 2002 c

ies on Libet’s findings. Consciousness and Cognition 12: 321–33
Libet, B., W. W. Alberts, E. W. Wright, L. Delattre, G. Levin, and

1964. Production of threshold levels of conscious sensation
stimulation of human somatosensory cortex. Journal of Neu
27:546–578.
Libet, B., W. W. Alberts, E. W. Wright, and B. Feinstein. 1967. Resp

human somatosensory cortex to stimuli below threshold fo
sensation. Science 158:1597–1600.
BIBLIOG
Libet, B., C. A. Gleason, E. W. Wright, and D. K. Pearl. 1983. Tim
scious intention to act in relation to onset of cerebral activ
ness-potential): the unconscious initiation of a freely volun
Brain 106:623–642.
Libet, B., E. W. Wright, Jr., B. Feinstein, and D. K. Pearl. 1979. Su

ral of the timing for a conscious sensory experience: a fun
for the somatosensory specific projection system in man.
224.
———. 1992. Retroactive enhancement of a skin sensation by a d

cal stimulus in man: evidence for delay of a conscious sen
ence. Consciousness and Cognition 1:367–375.
Libet, B., E. W. Wright, and C. Gleason. 1982. Readiness-potentia

unrestricted “spontaneous” vs. pre-planned voluntary acts
cephalography & Clinical Neurophysiology 54:322–335.
Marshall, J. C. 1989. An open mind? Nature 339:25–26.
Marshall, L. H., and H. W. Magoun. 1998. Discoveries in the Huma

Totowa, N.J.: Humana Press.
McGinn, C. 1997. Minds and Bodies: Philosophers and Their Ideas. L

ford University Press.
———. 1999. Can we ever understand consciousness? The New Yo

June 10, 1999, pp. 44–48.
Melchner, L. von, S. L. Pallas, and M. Sur. 2000. Visual behavior m

retinal projections directed to the auditory pathway. Natur
234 •
BIBLIOGRAPHY
cations.
Nichols, M. J., and W. T. Newsome. 1999. Monkeys play the odds.

400:217–218.
Nishimura, H. 1999. Visual stimuli activate auditory cortex in the

9:392–405.
Palm, A. 1991. Book review of The Emperor’s New Mind by R. Penr

Perspectives 2(1):27–28.
Penfield, W. 1958. The Excitable Cortex in Conscious Man. Liverpool:

University Press.
Penfield, W., and E. Boldrey. 1937. Somatic, motor and sensory rep
in the cerebral cortex of man as studied by electrical stimu
60:389–443.
Penfield, W., and H. Jasper. 1954. Epilepsy and the Functional Anatom
man Brain. Boston: Little, Brown and Company.
Penfield, W., and T. B. Rasmussen. 1950. The Cerebral Cortex of Ma

Macmillan Books.
Penrose, R. 1989. The Emperor’s New Mind: Concerning Computers, M

Laws of Physics. London: Oxford University Press.
Pepper, S. C. 1960. A neural-identity theory of mind. In: Dimension

ed. S. Hook. Washington Square: New York University Pre
Pieron, H., and J. Segal. 1939. Sur un phenomene de facilitation re

dan l’excitation electrique de branches nerveuses cutanées
tactile. Journal of Neurophysiology 2:178–191.
BIBLIOG
Poppel, E. 1988. Time and Conscious Experience. New York: Harcou
Jovanovich.
Popper, K. R. 1953. Language and the body-mind problem: a resta

interactivism. In: Proceedings of the XIth International Congr
phy, vol. 7. Amsterdam: North Holland Press, pp. 101–107.
———. 1992. In Search of a Better World: Lectures and Essays from T

London: Routledge.
Popper, K. R., and J. C. Eccles. 1977. The Self and Its Brain. Heidel
Springer-Verlag.
Popper, K. R., B. I. B. Lindahl, and P. Århem. 1993. A discussion o

body problem. Theoretical Medicine 14:167–180.
Ray, P. G., K. J. Meador, C. M. Epstein, D. W. Loring, and L. J. Da

netic stimulation of visual cortex: factors influencing the p
phosphenes. Journal of Clinical Neurophysiology 15(4):351–357
Ray, P. G., K. J. Meador, J. R. Smith, J. W. Wheless, M. Sittenfeld,

Clifton. 1999. Physiology of perception: cortical stimulatio
cording in humans. Neurology 52(2):1044–1049.
Roland, P. E., and L. Freiberg. 1985. Localization of cortical areas

thinking. Journal of Neurophysiology 53:1219–1243.
Rosenthal, D. M. 2002. The timing of conscious states. Consciousn

tion 11(2):215–220.
Schreiber, F. R. 1973 [1974]. Sybil, 2nd ed. New York: Regnery War
Schwartz, J., and S. Begley. 2002. The Mind and the Brain: Neuropla
Power of Mental Force. New York: Regan Books.
236 •
BIBLIOGRAPHY
———. 1993. Discussion in Libet, B. The neural time factor in con
unconscious events. In: Experimental and Theoretical Studies
ness. Ciba Foundation Symposium #174. Chichester, Englan
Wiley and Sons, p. 156.
———. 2000a. Consciousness, free action and the brain. Journal of

ness Studies 7(10):3–32.
———. 2000b. Consciousness. Annual Review of Neuroscience 2000 2
Sharma, J., A. Angelucci, and M. Sur. 2000. Visual behavior media

nal projections directed to the auditory pathway. Nature 40
Sherrington, C.S. 1940. Man on His Nature. Cambridge, England: C

University Press.
Shevrin, H. 1973. Brain wave correlates of subliminal stimulation,

attention, primary- and secondary-process thinking, and re
Psychological Issues 8(2); Monograph 30:56–87.
Shevrin, H., and S. Dickman. 1980. The psychological unconsciou

sary assumption for all psychological theory? American Psyc
35:421–434.
Singer, I. B. 1968 [1981]. Interview by H. Flender. In: Writers at Wo

Plimpton. New York: Penguin Books.
Singer, W. 1990. Search for coherence: a basic principle of cortical

zation. Concepts in Neuroscience 1:1–26.
———. 1991. Response synchronization of cortical neurons: an

epiphenomenon of a solution to the binding problem. IBRO
BIBLIOG
Kans.: University of Wichita.
Sokoloff, L., M. Reivich, C. Kennedy, M. H. Des Rosiers, C. S. Pa

Petttigrew, D. Sakurada, and M. Shinohara. 1977. The [14C]
deoxyglucose method for the measurement of local cereb
utilization; theory, procedure, and normal values in the co
anesthetized albino rat. Journal of Neurochemistry 28:897–91
Spence, S. A. 1996. Free will in the light of neuro-psychiatry. Philo

atry & Psychology 3:75–90.
Sperry, R. W. 1947. Cerebral regulation of motor coordination in

lowing multiple transection of sensorimotor cortex. Journ
Neurophysiology 10:275–294.
———. 1950. Neural basis of spontaneous optokinetic response p

visual inversion. Journal of Comparative and Physiological Psy
43:482–489.
———. 1952. Neurology and the mind-brain problem. American S

40:291–312.
———. 1980. Mind-brain interaction: mentalism, yes; dualism, no

ence 5:195–206.
———. 1984. Consciousness, personal identity and the divided br

Neuropsychologia 22:661–673.
———. 1985. Science and Moral Priority. Westport: Praeger.
Sperry, R. W., M. S Gazzaniga, and J. E. Bogen. 1969. Interhemisp

tionships: the neocortical commissures. Syndromes of hem
connection. In: Handbook of Clinical Neurology, eds. P. J. Vin
G. W. Bruyn. Amsterdam: North Holland Press, pp. 273–29
238 •
BIBLIOGRAPHY
Stoerig, P., A. Zantanon, and A. Cowey. 2002. Aware or unaware:
of critical blindness in four men and a monkey. Cerebral Cor
12(6):565–574.
Stratton, G. M. 1897. Vision without inversion of the retinal image

cal Review 4:341–360.
Taylor, J. L., and D. I. McCloskey. 1990. Triggering of pre-program

ments as reactions to masked stimuli. Journal of Neurophysio
446.
Vallbö, A.B., K. A. Olsson, K. G. Westberg and F. J. Clark. 1984.

Microstimulation of single tactile afferents from the human
107:727–749.
Velmans, M. 1991 Is human information processing conscious? Beh

Brain Sciences 14:651–669.
———. 1993. Discussion in Experimental and Theoretical Studies of C

Ciba Foundation Symposium #174, Chichester, England: Jo
and Sons, pp. 145–146.
Wegner, D. M. 2002. The Illusion of Conscious Will. Cambridge, Ma

Books (MIT Press).
Weiskrantz, L. 1986. Blindsight: A Case Study and Implications. Oxfo

don Press.
Whitehead, A. N. 1911. Quoted by Bruce Bower. 1999. Science New
———. 1925. Science and the Modern World. New York: McMillan.
Wieseltier, L. 1987. Book review of More Die of Heartbreak by S. B

New Republic, August 31, 1987, pp. 36–38.
BIBLIOG
of phenotypic variation in monozygotic twins by caudate
receptor binding. Science 273:1225–1227.
Wood, C. C., D. D. Spencer, T. Allison, G. McCarthy, P. P. William

W. R. Goff. 1988. Localization of human somatosensory c
surgery by cortical surface recording of somatosensory ev
tial. Journal of Neurosurgery 68(1):99–111.
Wouk, H. 1988. This Is My God. Boston: Little, Brown and Comp
240 •
BIBLIOGRAPHY
INDEX
“act now,” 126, 132, 135, 136, 141, 148, 149, “neuronal code,” 59; no
155, 161; sequence of events, 134 5, 14; mechanism for, 55
Adrian, Lord, 12 unique phenomenon, 13
alien hand sign, 129 awareness in volition, 132,
Amassian, V., 42 awareness of self, 4, 8, 13,
anesthesia, 50, 95, 204 axiomatic principles, 189
ascending sensory pathway, 42, 56, 67,
75, 90, 103, 106, 196 Baars, B. J., 166
athletes, 111 backward masking, 50, 51,
attention mechanism, 102, 115 backward referral, 64, 76,
awareness, 3, 15, 23, 28, 34, 38, 45, 46, 49, imental test of, 75
50, 55, 56, 57, 58, 59, 61, 62, 63, 64, 66, Barbur, 119, 120
67, 70, 71, 72, 75, 82, 88, 89, 90, 91, 93, baseball batter, 110, 111
96, 98, 101, 102, 104, 105, 106, 107, 109, Bellow, Saul, 214, 215
112, 116, 117, 120, 121, 124, 126, 128, 137, Berger, Hans, 24
141, 146, 163, 165, 166, 188, 189, 198, Bergson, Henri, 115
199, 200, 201, 205, 206, 210, 214, 222; Bernard, Claude, 31
all or nothing, 112; vs. content, 56; Berns, G. S., 93
vs. detection, 33; duration as Bishop Berkeley, 10
Boldrey, E., 26 87, 88, 158
Bower, Bruce, 99 computers, 131, 138, 141; a
brain areas for unconscious functions, self, 212, 213
119 conditional reflex, 27, 62,
brain death, 8 cortex, 27
Bremer, Frederic, 12 conscious content, 71, 72
Buchner, H., 85 conscious experience, un
Burns, B. D., 174 164, 165, 179
Burns, Jean, 157 Conscious mental field (C
Buser, P., 171 168, 169, 180, 207; conc
dualism, 181, 182; emer
carbon dioxide, 21 86, 163, 169, 182; neural
cardiac arrest, 216, 217, 218, 219; loss of 183; not physical, 168, 1
functions, 216, 218; reversibility, 217 176, 207
Carroll, Lewis, 4 consciousness and the br
Cartesian dualism, 86, 182, 183, 186, 203, lative relationships, 158
221 aspect theory, 158, 182;
central fissure, 43, 45 lem,” 158, 161; indepen
centrencephalic system, 19 158, 181, 182
cerebral stroke, 67, 81 conscious-unconscious, 9
Chalmers, David, 158, 159 conscious veto, 137, 143, 1
chemical transmitter, 97, 167 and unconscious origin
choices, deliberation, 132, 140, 155, 208 conscious will (W), 2, 122
Churchland, P. S., 18, 152, 159, 160 134, 136–139, 141, 142, 14
Churchland, Paul, 152, 159, 160 152, 154, 155, 161, 178, 18
Clark, R. E., 62, 63, 93 corpus callosum, 37
classical conditioning, 62, 64; simple de- Courtois, G., 174
lay conditioning, 51, 63, 64; trace Cowey, A., 15
conditioning, 63, 64 Crawford, B. H., 51
Cleckley, Harvey, 208 creativity, 5, 96
clock time, 98, 126, 127, 128, 130, 131, 133, Crick, Francis, 5
134 Cushing, Harvey, 26
242 •
INDEX
Death, 6, 8, 31, 214, 216, 219, 220 pulses, 45; of sensory co
Deecke, L., 124, 130 178; of “silent” areas, 27
delayed conditioning, 51 51; stimulus intensities,
delay in sensory awareness, 33, 45, 46, 56–58, 66, 73, 103, 116; su
50, 51, 55, 70, 71, 73, 75, 77, 86, 91, 199; pulses, 56, 106, 118; of te
half-second delay, 46, 50, 68, 70, 72, 177; train durations of p
75, 80, 88, 112, 123, 132, 199 45, 49, 57, 58; of visual a
deliberation, 132, 135, 148, 149, 155 cortex, 177
Dennett, Daniel, 18, 59, 64–66, 152, 159 electroconvulsive shock th
Descartes, R., 10, 79, 160, 181, 185, 186, electrophysiology, 23
187, 188, 193, 198 emergent phemonema, 16
Descartes, “thinking,” 186, 187, 188 emergent property, 86, 169
Descartes, not thinking, 188 endogenous conscious ele
determinism and free will, 6, 151, 154, logram (EEG), 24, 25, 17
156, 168, 198, 215, 216; as illusion, 144, functions, 89, 113, 128, 14
152; non-determinism, 152–154; scien- waves,” 24, 217
tific options, 156 ethical implications, 140, 1
determinist materialism, 5, 6 197; guilt, 150, 151, 201, 2
Dickman, S., 94, 207 control, 20, 149
discontinuity of conscious events, 112 event-related-potentials (E
Doctorow, E. L., 108 later components of, 48
Doty, Robert W., 133, 134, 139, 156, 162, evoked potentials (EPs), 47
165, 168, 171, 210 85, 113
dreaming, 14, 92, 96, 97 existentialist view, 72
dualism, 86, 181, 182, 183, 186, 194 eyelid blink reflex, 93
Eccles, Sir John, 12, 125, 165, 167, 168, fate, 140, 141
179, 204, 216 Feinstein, Bertram, 28, 29,
Echlin, F. A., 174, 176 ferrets, 84
Eddington, Sir Arthur, 162 Feynman, Richard, 9
Edelman, G. M., 166 Foerster, O., 108
Eimer, M., 135 forced choice, 15, 111, 116, 1
129; initiation by brain, 123; uncon- identity theory, 12, 13, 87, 14
scious initiation, 142, 144 imagination of moving, 1
free will, 4, 123, 125, 133, 136, 139, 141, 143, informed consent, 29, 30,
144, 145, 146, 149, 151, 152, 154–156, 161, Ingvar, David, 21, 22, 172,
166, 167, 197, 198, 201, 212, 215, 222; initiating an act, experien
control function, 146, 222; do we 123, 124, 136, 137, 139, 14
have it?, 123, 140; Kabbalist view, 141; 202, 222
trigger function, 142, 145, 197; uncon- inner quality, 12, 87, 159
scious acts, 140; unconscious initia- intention to act, 98, 109, 1
tion, 140, 145 142; awareness of, 98, 1
Freud, S., 2, 71, 120, 122, 200 introspective reports, 8, 9
Friberg, L., 22, 179 18, 29, 59, 85, 158, 169, 1
functionalism, 11 haviorism, 11, 17; non-v
objective evidence, 11; r
Gerard, R. W., 29 intuitive feeling, 94, 155
“ghost in machine,” 183, 221 involuntary acts, 129, 142
Goff, W. R., 69 isolated slab of cortex, 17
Golden Rules, 150 condition and, 176–177;
Goldring, S. J., 174 176–177; Sperry techniq
Granit, Ragnar, 12
Gray, C. M., 166 James, William, 1, 112
Grossman, R., 42 Japenga, Ann, 220
guessing responses, 104, 105, 106, 117 Jasper, Herbert, 12, 18, 19,
Jensen, Arthur, 54, 55
Haggard, P., 135 Josephson, Brian D., 162,
Halliday, A. M., 51 Jung, R., 108
Hawking, S., 183
Heckhausen, H., 135 Keller, I., 135
hippocampus, 26, 60, 61, 62, 63, 64; Kihlstrom, J. R., 94, 100
bilateral loss of, 61, 62 Koch, Christof, 4
Houdini, H., 220 Koestler, A., 97
human subjects, 9, 15, 22, 27, 60, 117, Kornhuber, H. H., 124, 13
129, 176, 177, 222 Kristiansen, K., 174
244 •
INDEX
Leksell stereotaxic frame, 28 mind, as loose cannon, 183
lesions in brain, 15, 19, 20, 30, 63, 204, mind, not divisible, 181, 19
205, 206, 207 mind and brain, separation
Libet, B., 30, 38, 40, 48, 49, 50, 52, 56, 58, mind-brain relation, 7, 16,
59, 64, 65, 66, 67, 73, 77, 85, 98, 102, 146, 160, 161, 164, 166, 16
113, 126, 127, 131, 135, 137, 145, 147, 155, 222
158, 161, 166, 168, 169, 170, 177, 201, Mingay, R., 51
202, 213 models or theories, 3, 17, 1
Libet, Fay, 98 motor units, 114, 115
Loewi, O., 96, 97 Mountcastle, V., 164, 166
Lundberg, Anders, 12 Mt. Zion Hospital, San Fr
Mukhopadhyay, A. K., 181
magnetic resonance imaging (MRI), 22, multiple personality disord
23, 179; neural changes in time, 23; 205, 208
resolution of time, 23
magnetoencephalogram (MEG), 25 Nagel, Thomas, 12
maps of cortex, 36 near-death experiences, 21
Marshall, John C., 31 219; experimental tests o
Marshall, L. and H. W. Magoun, 4 neuronal representations,
materialism, eliminative, 5, 6, 159 neuronal time, 72, 122
McGinn, Colin, 159, 160, 183 Newsome, W. T., 94
Meador, K., 42, 43 Newton, 17
medial lemniscus, 43, 44, 45, 56, 75, 76, Nichols, M. J., 94
78, 147
medulla oblongata, 43 objective scientific study, 1
Melchner, L. von, 84 obsessive-compulsive diso
memory, 15, 26, 54, 59, 60, 61, 62, 64, 65, 142, 143
66, 94, 180, 205, 209, 218; explicit (de- outer quality, 87, 159
clarative), 60, 61, 64; implicit (non- overlapping events, 114
declarative), 61, 62; role in aware-
ness, 59, 60, 63, 64, 66, 92 Palm, Ann, 212
mental events, discontinuous, 114 paradoxical timings, 75
mental influence, neuronal, 166, 167, Penfield, Wilder, 129, 177
amnesia and, 205; and cerebral hemi-
sphere loss, 205; “split brain,” 210 qualia, 14
Phillips, Charles, 12
Phillipson, L., 179 Rasmussen, T. B., 37, 129
Phineas Gage, 20 Ray, P. G., 42
physical vs. non-physical activities, 3, reaction times (RT), 54, 5
158, 160, 161, 168, 221 deliberate lengthening
physical vs. subjective, 2, 3, 5, 9, 14, 27, readiness potential (RP),
32, 153, 157, 158, 159, 162, 163, 168, 169, 126, 127, 129, 130, 131, 13
179, 183, 184, 221, 222; unexplained 136, 137, 138, 139, 141, 14
gap between, 153; and violation of reductionism, 215, 216
laws, 154 regional cerebral blood fl
pineal gland, 181, 190, 191 22; in motor activity, 21
Pippard, Sir Brian, 216 processes, 21
pitched ball, timings, 110 responsibility, 140, 202, 20
playing musical instruments, 97, 109, retinal pathway, 84
200 retroactive enhancement,
Poincaré, H., 95, 96, 195 right hemisphere, 37, 67,
Poppel, E., 198 conscious?, 171
Popper, Karl, 3, 6, 12, 125, 152, 165, 169, Roland, P. E., 179
189, 190, 204, 212, 216 Rolandic fissure, 26, 36, 37
positron emission tomography (PET), Rosenthal, David M., 201
22, 23, 179 RPI, 132, 133, 134, 135
postcentral gyrus, 26, 40, 44, 83, 177 RPII, 132, 133, 134, 135, 136,
postsynaptic responses, 24 Russell, Bertrand, 108
preceding voluntary act, 131, 141
precentral gyrus, 26 Schreiber, Flora R., 209
preplanning, 129, 130, 131, 132, 133, 136, Schwartz, J. M., 143, 220
137, 148, 149, 155 Searle, John, 99, 100, 135,
preset timing, 129 161, 212
primary EP, 47, 48, 49, 50, 67, 74, 75, 76, self, 1, 8, 13, 136, 161, 171, 2
78, 80, 85, 86; function of, 67; sensa- 206, 207, 208, 209, 210,
246 •
INDEX
self, singular, 208 Spinoza, B., 159, 215
self and soul, 6, 7, 191, 192, 194, 203, split-brain patients, 171, 20
204, 211, 213, 214, 215, 216, 220, 221 Squire, L. R., 62, 63, 93
self-awareness, 13, 86 Stoerig, Petra, 15
self-initiated acts, 131, 137 stream of consciousness, 1
self-paced acts, 130 subjective experience, 2, 3,
selves one or two, 171, 211 16–18, 27–29, 32, 82, 84,
sensory awareness delayed. See delay in 100, 121, 137, 153, 157–159
sensory awareness 168, 169, 172, 180, 183, 18
sensory modality vs. cortical area, 68, 201, 216, 218, 221, 222; ac
164 individual only, 1, 8, 87,
sensory pathways, 48, 51, 68 169, 182, 183; arising from
sensory perception, at brain, 116, 117 157, 159, 163, 168; fundam
Sharma, J., 84 erty, 163
Sherrington, Sir Charles, 165 subjective inner life, 1, 2, 2
Shevrin, H., 12, 94, 118, 120, 207 subjective referral, 75, 77, 7
signal detection, 111, 116 as neural corrector, 81;
silent cortex, 27 sphere?, 186; neuronal m
silicon chip brain, 213 85
sin, 150, 152 subjective referral of sensa
Singer, Isaac Bashevis, 156, 166 86; antedating, 68, 72, 7
Singer, W., 166 91, 98, 113, 132, 199, 200;
singing, 109 referral in time, 78, 87,
single-case studies, 31 199; spatial, 80, 87; temp
single skin pulse, 46, 50; delay for subjective timing, 64, 67, 6
awareness, 46, 53–55 78, 79, 81, 85, 113, 137
Sokoloff, Louis, 22 subliminal perception, 106
soul, 6, 7, 191, 192, 194, 203, 204, 211, 213, subthreshold sensory resp
214, 215, 216, 220, 221 supplementary motor area
speaking, 108, 148, 171, 210 132, 178
specific projection pathway, 45, 47, 81 switching tasks, 133
speculations, 3, 113, 166, 204 Sybil story, 209
32, 34, 71, 89, 90, 91, 92,
Tasker, 42 99, 100, 101, 102, 106, 11
testable hypotheses, 3, 17, 32, 55, 102, 146, 147, 148, 149, 196, 1
136, 159, 161, 168, 172, 182 201, 207, 208; distinctio
thalamus, 14, 19, 43, 44, 45, 49, 69, 103, scious, 2, 8, 16, 38, 90, 9
207; ventrobasal, 43 106, 117, 118, 119, 122, 14
Thompson, Richard, 93 unconscious solutions, 95
thought processes, 21, 95, 112
threshold sensory experience, 38, 73; Velmans, Max, 66, 92, 145
liminal intensity of stimulus, 39, 40, visual image, subjective r
41, 42, 56, 57, 58, 73 visual image upside down
timelessness, 88 visuomotor referral, 82
“time-on,” filter function, 115 voluntary acts, 93, 123, 125
“time-on,” test, 102–106, 116, 118 141, 144, 148, 149, 151, 20
time-on theory, 92, 101, 102, 103, 106,
107, 108, 109, 111, 112, 115, 116, 118, 120, Wegner, D., 144, 152, 156
198, 199 Weiskrantz, L., 117, 196
timing “clock,” 126, 127, 128, 130, 131, Whitehead, A. N., 95, 98
132, 133, 134 Wieseltier, L., 214, 215
timing of skin stimulus, 128, 131, 132, Wigner, Eugene, 6, 157
133, 137 Wilbur, Cornelia, 209
Tourette’s syndrome, 129, 142, 143, 146 Wittgenstein, Ludwig, 88
Turing paradigm, 212, 213 writing, 97, 98, 109
unconscious detection, 16, 101, 103, 106, yourself, 4, 140, 196, 198
116, 117, 118; little delay, 106, 107, 117;
reaction time (RT), 54, 111, 136, 199
248 •
INDEX

(Perspectives in Cognitive Neuroscience) Benjamin Libet - Mind Time - The Temporal Factor in Consciousness (2005, Harvard University Press)

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

(Perspectives in Cognitive Neuroscience) Benjamin Libet - Mind Time - The Temporal Factor in Consciousness (2005, Harvard University Press)

Enviado por

Direitos autorais:

Formatos disponíveis

PERSPECTIVES IN COGNITIVE NEUR

Printed in the United Sta

Library of Congress Cataloging-in-Pu

1. Introduction to the Question

I just typed the word “consciousness” into the sear

Klein, S. 2002. Libet’s research on the timing of conscious intentio

Wegner, D. M. 2002. The Illusion of Conscious Will. Cambridge, M

How did I come to write this book?

Something happens when to a certain brain sta

—William James (1899)

You stop to admire the intense blue of a flower; you

The Problem: Relating Brain

General Views on Mind and Matter

Is There Any Scientific

How Can We Study the Relation between the

1. Introspective report as the operational criterion. I h

2. No a priori rules for mind-brain relationship. A coro

Where in the Brain Are the

Our Experimental Entry

If you tap your finger on a table, you experience the e

Initial Evidence from Cerebral Stimulation

[To view this image, refer to

Fig. 2.1. Maps of human cerebral cortex.

medial side of the right hemisphere. The frontal region is to

Fig. 2.2. Duration of stimulus train of pulses, relative to prod

stimulus pulses, if the liminal intensity for a given fr

Fig. 2.3. Pathways for the touch, pressure, position-sense sens

Actual Delay in Awareness with Normal Senso

Fig. 2.4. Electrical responses (event-related-potentials, ERP) of

does not elicit any subjective sensation at all. This i

applied repetitively here to produce a conscious sens

Retroactive, Backward Effects

[To view this image, refer to

Fig. 2.6. Retroactive effects of a delayed cortical stimulus that

Efforts to Deliberately Slow Responses

How Does the 0.5-Sec

Fig. 2.7. Direct cortical responses (DCRs) evoked in SI cortex b

pps was significant. It was the duration of train that

Role of Memory Formation

Explicit Memory and the Hippocampus Structu

Classical Conditioning and Awareness

[a] conjoint operation of the hippocampal system

An important implication of this finding is that trace

Other Evidence on Proposals of Memory for Aw

Conscious synchronicity of sensations. This brings up

Why Get Excited about the Delay in Sensory A

Antedating of Delayed Sensory Experience

Fig. 2.8. Subjective timing of sensory experience with cerebral ve

tually produced the experience). Indeed, we ran an e

Fig. 2.9. Diagram of hypothesis for subjective referral of a sensor

Subjective Referrals “Correct” the Neural Repre

General Implications of Subjective Referrals

Neural Mechanisms for Subjective Referrals?

You are driving along in your car at 30 mph on a city

Unconscious Mental Functions

Fig. 3.1. Sequence of events when a boy steps in front of a movin

Are Unconscious Functions “Mental”?

Experimental test of time-on theory. Any theory prop

Fig. 3.3. Timings for baseball batter’s response to a pitched ba

into this category. A professional tennis player must re

Fig. 3.4. Overlap of discontinuous mental events and the fee

thoughts is explainable by an overlapping of the differ

How the brain deals with voluntary acts is an issue