ArticlePDF AvailableLiterature Review

A Framework for Consciousness



Here we summarize our present approach to the problem of consciousness. After an introduction outlining our general strategy, we describe what is meant by the term 'framework' and set it out under ten headings. This framework offers a coherent scheme for explaining the neural correlates of (visual) consciousness in terms of competing cellular assemblies. Most of the ideas we favor have been suggested before, but their combination is original. We also outline some general experimental approaches to the problem and, finally, acknowledge some relevant aspects of the brain that have been left out of the proposed framework.
nature neuroscience • volume 6 no 2 • february 2003 119
General strategy
The most difficult aspect of consciousness is
the so-called ‘hard problem’ of qualia1,2
the redness of red, the painfulness of pain,
and so on. No one has produced any plau-
sible explanation as to how the experience
of the redness of red could arise from the
actions of the brain. It appears fruitless to
approach this problem head-on. Instead,
we are attempting to find the neural corre-
late(s) of consciousness (NCC), in the hope
that when we can explain the NCC in causal
terms, this will make the problem of qualia
clearer3. In round terms, the NCC is the
minimal set of neuronal events that gives
rise to a specific aspect of a conscious per-
cept. We discuss elsewhere4why we think
consciousness has to be largely private. By
‘private’, we mean that it is accessible exclu-
sively to the owner of the brain; it is impos-
sible for me to convey to you the exact
nature of my conscious percept of the color
red, though I can convey information about
it, such as whether two shades of red appear
to me to be the same or different.
Our main interest is not the enabling
factors needed for all forms of conscious-
ness, such as the activity of the ascending
reticular systems in the brainstem. Rather,
we are interested in the general nature of
the neural activities that produce each
particular aspect of consciousness, such
as perceiving the specific color, shape or
movement of an object.
As a matter of tactics, we have con-
centrated on the visual system of primates,
leaving on one side some of the more dif-
ficult aspects of consciousness, such as
emotion and self-consciousness. Our
see, for example, either introns or RNA
editing. And who would have guessed
that DNA replication usually starts with
the synthesis of a short stretch of RNA,
which is then removed and replaced by
DNA? The broad framework acted as a
guide, but careful experimentation was
needed for the true details to be discov-
ered. This lesson is broadly applicable
throughout biology.
A preamble on the cerebral cortex
By the ‘cortical system’, we mean the cere-
bral cortex plus other regions closely asso-
ciated with it, such as the thalamus and
the claustrum, and probably the basal
ganglia, the cerebellum and the many
widespread brainstem projection systems.
We shall refer to the ‘front’ of the cor-
tex and the ‘back’ of the cortex because
the terms ‘frontal’ and ‘prefrontal’ can be
ambiguous. For example, is the anterior
cingulate prefrontal?
The dividing line between front and
back is somewhat arbitrary. It roughly
coincides with the central sulcus. It may
turn out that a good operational defini-
tion is that the front is all those parts that
receive a significant input, via the thala-
mus, from the basal ganglia. (This simple
division is probably not useful for olfac-
tion, however.)
There is an absolutely astonishing vari-
ety and specificity of actions performed
by the cortical system. The visual system
of the higher mammals handles an almost
infinite variety of visual inputs and reacts
to them in detail and with remarkable
accuracy. It is clear that the system is high-
ly evolved, is likely to be specified epige-
netically in considerable detail and can
learn a large amount from experience.
The main function of the sensory cor-
tex is to construct and use highly specific
feature detectors, such as those for orien-
tation, motion or faces. The features to
which any cortical neuron responds
A framework for consciousness
Francis Crick and Christof Koch
Here we summarize our present approach to the problem of consciousness. After an
introduction outlining our general strategy, we describe what is meant by the term
‘framework’ and set it out under ten headings. This framework offers a coherent scheme for
explaining the neural correlates of (visual) consciousness in terms of competing cellular
assemblies. Most of the ideas we favor have been suggested before, but their combination is
original. We also outline some general experimental approaches to the problem and, finally,
acknowledge some relevant aspects of the brain that have been left out of the proposed
framework may well apply, however, to
other sensory modalities. We have been
especially interested in the alert macaque
monkey because to find the NCC, not
only widespread neural activities but also
the detailed behavior of single neurons (or
small groups of neurons) must be inves-
tigated on very fast time scales. This is dif-
ficult to do systematically in humans.
Methods such as fMRI are too coarse in
both space and time to be of much use for
our problem. On the other hand, experi-
ments in visual psychology are much eas-
ier to do on humans than on monkeys.
Moreover, humans can report what they
are conscious of and thus convey the ‘con-
tents’ of their consciousness. For these
reasons, experiments on monkeys and
humans should be pursued in parallel.
A framework is not a detailed hypothesis
or set of hypotheses; rather, it is a sug-
gested point of view for an attack on a sci-
entific problem, often suggesting testable
hypotheses. Biological frameworks differ
from frameworks in physics and chem-
istry because of the nature of evolution.
Biological systems do not have rigid laws,
as physics has. Evolution produces mech-
anisms, and often sub-mechanisms, so
that there are few ‘rules’ in biology which
do not have occasional exceptions.
A good framework is one that sounds
reasonably plausible relative to available
scientific data and that turns out to be
largely correct. It is unlikely to be correct
in all the details. A framework often con-
tains unstated (and often unrecognized)
assumptions, but this is unavoidable.
An example from molecular biology
might be helpful. The double-helical
structure of DNA immediately suggest-
ed, in a novel way, the general nature of
gene composition, gene replication and
gene action. This framework turned out
to be broadly correct, but it did not fore-
Francis Crick is at the Salk Institute for
Biological Studies, 10010 N. Torrey Pines Road,
La Jolla, California 92037, USA.
Christof Koch is at the California Institute of
Technolog y, 1200 East California Boulevard,
Pasadena, California 91125, USA.
© 2003 Nature Publishing Group
120 nature neuroscience • volume 6 no 2 • february 2003
are usually highly specific but multi-
dimensional. That is, one neuron does not
respond to a single feature but to a family
of related features. Such features are some-
times called the ‘receptive field’ of that
neuron—the ‘non-classical receptive
field’5expresses the relevant context of the
‘classical receptive field’. The visual fields
of neurons higher in the visual hierarchy
are larger and respond to more complex
features than those that are lower down.
An important but neglected aspect of
the firing of a neuron (or a small group of
associated neurons) is its ‘projective
field’6. This term describes the perceptual
and behavioral consequences of stimulat-
ing such a neuron in an appropriate man-
ner (for further discussion of motor and
premotor cortex, see ref. 7). Both the
receptive field and the projective field are
dynamic, not merely static, and both can
be modified by experience.
How are feature detectors formed? A
broad answer is that neurons do this by
detecting common and significant corre-
lations in their inputs and by altering their
synapses (and perhaps other properties)
so that they can more easily respond to
such inputs. In other words, the brain is
very good at detecting apparent causation.
Exactly how it does this is more contro-
versial. The main mechanism is probably
Hebbian, but Hebb’s seminal suggestion
needs to be expanded.
All the above might suggest that cor-
tical action is highly local. Nothing could
be further from the truth. In the cortex,
there is continual and extensive interac-
tion, both among neighboring cells and
also very widely, thanks to the many long
cortico-cortical and cortico-thalamo-cor-
tical routes. This is much less true of the
thalamus itself.
The sensory cortex is arranged in a
semi-hierarchical manner. (This is cer-
tainly true of the visual cortex8, but is less
clear in the front of the brain.) That is,
most cortical areas do not detect simple
correlations in the sensory input, but
detect correlations among correlations
being expressed by other cortical areas.
This remarkable feature of the cortex is
seldom emphasized.
There has been a great selective advan-
tage in reacting very rapidly, for both
predators and prey. For this reason, the
best is the enemy of the good. Typically, it
is better to achieve a rapid but occasion-
ally imperfect performance instead of a
more prolonged one that always produces
a perfect result. Another general principle
may be to use several rough and ready
methods in parallel to reach a conclusion,
(Further discussion of this idea comes at
the end of subhead #3 below.)
We have discussed elsewhere17 whether
the neural activity in the front of the brain
is largely unconscious. One proposal18,19,
for example, is that humans are not direct-
ly conscious of their thoughts, but only of
sensory representations of them in their
imagination. At the moment, there is no
consensus about this20.
The hypothesis of the homunculus
is very much out of fashion these days,
but this is, after all, how everyone
thinks of themselves. It would be sur-
prising if this overwhelming illusion
did not reflect in some way the general
organization of the brain.
2. Zombie modes and consciousness
Many actions in response to sensory
inputs are rapid, transient, stereotyped
and unconscious14,21. They could be
thought of as cortical reflexes. Con-
sciousness deals more slowly with broad-
er, less stereotyped aspects of the sensory
inputs (or a reflection of these in imagery)
and takes time to decide on appropriate
thoughts and responses. It is needed
because otherwise, a vast number of dif-
ferent zombie modes would be required.
The conscious system may interfere some-
what with the concurrent zombie system.
It seems to be a great evolutionary advan-
tage to have zombie modes that respond
rapidly, in a stereotyped manner, togeth-
er with a slightly slower system that allows
time for thinking and planning more
complex behavior.
Visual zombie modes in the cortex
probably use the dorsal stream in the pari-
etal region14. Some parietal activity, how-
ever, also affects consciousness by
producing attentional effects on the ven-
tral stream, at least under some circum-
stances. The conscious mode for vision
depends largely on the early visual areas
(beyond V1) and especially on the ventral
stream. There are no recorded cases of
purely parietal damage which led to a
complete loss of conscious vision.
In a zombie mode, the main flow of
information is probably feed-forward. It
could be considered a forward-traveling
net-wave. A net-wave is a propagating
wave of neural activity, but it is not the
same as a wave in a continuous medium.
Neural networks in a cortex have both
short and long connections, so a net-wave
may, in some cases, jump over interven-
ing regions. In the conscious mode, it
seems likely that the flow is in both direc-
tions (see #5 below) so that it resembles
more of a standing net-wave10.
rather than following just one method
very accurately. This appears to be how,
for example, people see in depth.
Incoming visual information is often
incomplete or ambiguous. If two similar
stimuli are presented in rapid succession,
the brain blends them together into one
percept. If they are different but in con-
tradiction, such as a face and a house, the
brain selects one at a time (as in binocular
rivalry) instead of blending them togeth-
er. In cases where there is not enough
information to lead to an unambiguous
interpretation of one’s environment9, the
cortical networks ‘fill in’—that is, they
make their best guess, given the incom-
plete information. Such filling-in is like-
ly to happen in many places in the brain.
This general principle is an important
guide to much of human behavior (as in
‘jumping to conclusions’).
Our present framework
Having outlined a few general points
about the cortical system, let us now con-
sider specifically the NCC and its atten-
dant properties. We are mainly interested
in time periods on the order of a few hun-
dred milliseconds, or at the most, several
seconds, so we can now leave on one side
processes that take more time, such as
permanently laying down a new memo-
ry. We have listed the main ingredients of
our framework under ten headings. We
have not previously discussed in print
items 3, 5, 7 and 10, though we have men-
tioned the first three in a book chapter
that is still in press10. Many of our basic
ideas on consciousness, such as the
importance of attention and correlated
firing, were outlined in our 1990 paper11,
and in 1995, we suggested a plausible
function for consciousness12 and later
updated our ideas in a 1998 review3. Pre-
viously we proposed that people are not
directly conscious of the neural activity in
primary visual cortex (V1) and that the
(visual) cortex appears hierarchical
because it contains no strong loops12,13.
More recently, we have supported the sug-
gestion14 that in addition to a slower, all-
purpose conscious mode, the brain has
many ‘zombie modes’15, which are char-
acterized by rapid and somewhat stereo-
typed responses.
1. The (unconscious?) homunculus
A good way to begin to consider the overall
behavior of the cerebral cortex is to imag-
ine that that the front of the brain is ‘look-
ing at’ the sensory systems, most of which
are at the back of the brain. This division of
labor does not lead to an infinite regress16.
© 2003 Nature Publishing Group
nature neuroscience • volume 6 no 2 • february 2003 121
3. Coalitions of neurons
The cortex is a very highly and specifically
interconnected neural network. It has many
types of excitatory and inhibitory interneu-
rons and acts by forming transient coali-
tions of neurons, the members of which
support one another. ‘Coalitions’ implies
assemblies’—an idea which goes back at
least to Hebb22—plus competition among
them (see also ref. 23). On the basis of
experimental results in the macaque, some
researchers suggest that selective attention
biases the competition among rivalrous cell
assemblies, but they do not explicitly relate
this idea to consciousness24.
The various neurons in a coalition in
some sense support one another, either
directly or indirectly, by increasing the
activity of their fellow members. The
dynamics of coalitions are not simple. In
general, at any moment the winning coali-
tion is somewhat sustained, and embodies
what we are conscious of.
It may help to make a crude political
analogy. The primaries and the early
events in an election would correspond
roughly to the preliminary unconscious
processing. The winning coalition associ-
ated with an object or event would corre-
spond to the winning party, which would
remain in power for some time and would
attempt to influence and control future
events. ‘Attention’ would correspond to the
efforts of journalists, pollsters and others
to focus on certain issues rather than oth-
ers, and thus attempt to bias the electorate
in their favor. Perhaps those large pyrami-
dal cells in cortical layer 5 that project to
the superior colliculus and the thalamus
(both involved in attention) would corre-
spond to electoral polls. These progress
from early, tentative polls to later, rather
more accurate ones as the election
approaches. It is unlikely that all this hap-
pens in the brain in a fixed time sequence.
The brain may resemble more the British
system, in which the time between one
election and the next can be irregular. Such
an analogy should not be pressed too far.
Like all analogies, it should be regarded as
a possible source of ideas, which of course
will have to be confirmed by experiment.
Coalitions can vary both in size and in
character. For example, a coalition pro-
duced by visual imagination (with one’s
eyes closed) may be less widespread than a
coalition produced by a vivid and sus-
tained visual input from the environment.
In particular, representations of imagined
visions may fail to reach down to the
lower echelons of the visual hierarchy.
Coalitions in dreams may be somewhat
different from waking ones.
conscious unless there is an essential node
for it. For consciousness, there may be
other necessary conditions, such as pro-
jecting to the front of the brain12.
A node, all by itself, cannot produce
consciousness. Even if the neurons in that
node were firing appropriately, this would
produce little effect if their output synaps-
es were inactivated. A node is a node, not
a network. Thus a particular coalition is
an active network, consisting of the rele-
vant set of interacting nodes that tem-
porarily sustains itself.
Much useful information can be
obtained from lesions. In humans, the
damaged area is usually fairly large. It is
not clear what effects a very small, (possi-
bly bilateral) reversible lesion would have
in the macaque, as it is difficult to discov-
er exactly what a monkey is conscious of.
The smallest useful node may be a corti-
cal column29 or, perhaps, a portion of a
cortical column. The feature which that
node represents is (broadly) its columnar
property. This is because although a sin-
gle type of pyramidal cell usually sends its
information to only one or two cortical
areas, the pyramidal cells in a column pro-
ject the columnar property collectively to
many cortical areas, and can thus strength-
en any coalition that is forming.
5. The higher levels first
For a new visual input, the neural activity
first travels rapidly and unconsciously up
the visual hierarchy to a high level, possi-
bly in the front of the brain (this might
instantiate a zombie mode). Signals then
start to move backward down the hierar-
chy so that the first stages to reach con-
sciousness are at the higher levels
(showing the gist of the scene30,31; see also
ref. 32), which send these ‘conscious’ sig-
nals again to prefrontal cortex, followed
by corresponding activity at successive
lower levels (to provide the visual details).
This is an oversimplified description.
There are also many horizontal connec-
tions in the hierarchy.
How far up the hierarchy the initial
net-wave travels may depend upon
whether attention is diffused or focused
at some particular level.
6. Driving and modulating connections
In considering the physiology of coali-
tions, it is especially important to under-
stand the nature of neural connections.
The classification of neuronal inputs is
still in a primitive state. It is a mistake to
think of all excitatory neural connections
as the same type. First, connections to a
cortical neuron fall roughly into two
If there are coalitions in the front of
the cortex, they may have a somewhat dif-
ferent character from those formed at the
back of the cortex. There may be more
than one coalition that achieves winning
status and hence produces conscious
experience. The coalitions in the front
may reflect feelings such as happiness and,
perhaps, the feeling of ‘authorship’, which
is related to free will25. Such feelings may
be more diffuse and persist for a longer
time than coalitions in the back of cortex.
The terms ‘affect’ and ‘valuations’ are now
being used for what we have traditional-
ly called ‘feelings’18,19. Our first working
assumption (the homunculus) implies
that it is better not to regard the back plus
the front as one single coalition, but rather
as two or more separate coalitions that
interact massively, but not in an exactly
reciprocal manner.
4. Explicit representations
An explicit representation of a particular
aspect of the visual scene implies that a
small set of neurons exists that responds
as a detector for that feature, without fur-
ther complex neural processing. A possi-
ble probe, or an operational test, for an
explicit representation might be whether a
single layer of ‘neurons’ could deliver the
correct answer. For example, if a single-
layered neural network were fed the activ-
ity of retinal neurons, it would not be able
to recognize a face. Fed from the relevant
parts of inferior temporal cortex, howev-
er, it could reliably signal ‘face’ or ‘no face’.
There is much evidence from both
humans and monkeys that if there are no
such explicit neurons, or if they are all lost
by brain damage, then the subject is
unable to consciously perceive that aspect
directly. Well-known clinical examples are
achromatopsia (loss of color perception),
prosopagnosia (loss of face recognition)
or akinetopsia (loss of motion percep-
tion). In all cases, one or a few attributes
of conscious experience have been lost,
while most other aspects remain intact. In
the macaque, a small, irreversible lesion
of the motion area MT/V5 leads to a tem-
poral deficit in motion perception that
recovers within days. Larger lesions cause
a more permanent loss26.
Note that an explicit representation is
a necessary but not sufficient condition
for the NCC to occur.
One can describe this in terms of
essential nodes’27,28. The cortical neural
networks (at least for perception) can be
thought of as having nodes. Each node is
needed to express one aspect of one per-
cept or another. An aspect cannot become
© 2003 Nature Publishing Group
122 nature neuroscience • volume 6 no 2 • february 2003
broad classes: driving and modulating
inputs13. For cortical pyramidal cells, dri-
ving inputs may largely contact the basal
dendrites, whereas modulatory inputs
include back-projections (largely to the
apical dendrites) or diffuse projections,
especially those from the intralaminar
nuclei of the thalamus.
This classification may be too simple.
In some cases, a single type of input to a
neuron may be driving, such as the input
from the lateral geniculate nucleus (LGN)
to V1. In other cases, several types of ‘dri-
ving’ inputs may be needed to make that
neuron fire at a significant rate. It is pos-
sible that the connections from the back
of the brain to the front are largely driving,
whereas the reverse pathways are largely
modulatory, but this is not experimental-
ly established. This general pattern would
not hold for cross-modal connections.
The above tentative classification is
largely for excitatory cells. Strong loops
of driving connections probably do not
occur under normal conditions13. It
seems likely that cortical layer 5 cells
which project to the thalamus are dri-
ving, and that those from layer 6 are
7. Snapshots
Has a successful coalition any special char-
We propose that conscious awareness
(for vision) is a series of static snapshots,
with motion ‘painted’ on them34,35 (Fig. 1).
By this we mean that perception occurs in
discrete epochs. It is well established that
should be the underlying NCC (for exam-
ple, firing at either a low or high level).
This activity may also show hysteresis; that
is, it may stay there longer than its sup-
port warrants.
What could be special about this activ-
ity that reaches above the consciousness
threshold? It might be some particular
way of firing, such as a sustained high
rate, some sort of synchronized firing or
firing in bursts. Or it might be the firing
of special types of neurons, such as those
pyramidal cells that project to the front of
the brain39 (Fig. 2). This may seem
unlikely but, if true, would greatly sim-
plify the problem, both experimentally
and theoretically.
What is required to maintain this spe-
cial activity above the threshold? It might
be something special about the internal
dynamics of the neuron, perhaps involv-
ing the accumulation of chemicals such as
Ca2+, either in the neuron itself or in one
of its associated inhibitory neurons. It
might also be re-entrant circuits in the
cortical system40. Positive feedback loops
could, by iteratively exciting the neuron,
push its activity increasingly upward so
that the activity not only reaches above
some critical threshold, but is maintained
there for some time.
There are a few complications: the
threshold level may depend on the rate
of approach to the threshold or on how
long the input is sustained, or both of
these. At the beginning of each new
snapshot, there may be some hold-over
from the previous one.
Put another way, these are partial
descriptions of conscious coalitions form-
ing, growing or disappearing.
There is no evidence for a regular
clock in the brain on the second or frac-
tion-of-a-second time scale. The dura-
tion of any snapshot (or fragment of a
snapshot) is likely to vary somewhat,
depending on such factors as sudden on-
signals, off-signals, competition, habit-
uation and so on. Several psychological
effects have been described41, such as an
illusion under constant illumination
similar to the wagon-wheel effect, which
hint that there are some irregular batch-
like effects in vision.
the mechanisms for position-estimation
and for detecting motion are largely sepa-
rate. (Recall the motion after-effect.) Thus,
a particular motion can be represented by
a constant rate of firing of the relevant neu-
rons. It is not surprising, then, that if the
eyes do not move in smooth pursuit, the
brain is very poor at recognizing accelera-
tion even though it is good at distinguishing
movements36. (As perceived motion is con-
stant during a snapshot, it can only change
between snapshots, which suggests that
there is little or no explicit representation
for such a change. Hence, acceleration is
not easily seen). All the other conscious
attributes of the percept at that moment are
part of the snapshot.
The durations of successive snapshots
are unlikely to be constant. (They are dif-
ficult to measure directly.) Moreover, the
time of a snapshot for shape, say, may not
exactly coincide with that for, say, color.
It is possible that these durations may be
related to the αrhythm37 or even the δ
rhythm. The theory of the ‘perceptual
moment’ was suggested as early as 1955,
when it was not known how motion is
represented in the brain38, but in recent
years, it has been largely forgotten.
To reach consciousness, some (unspec-
ified) neural activity for that feature has
to cross a threshold. It is unlikely to do so
unless it is, or is becoming, the member
of a successful coalition. It is held above
threshold, possibly as a constant value of
the activity, for a certain time (the time of
that snapshot). As specific attributes of
conscious perception are all-or-none, so
Fig. 1. The snapshot hypothesis proposes that
the conscious perception of motion is not rep-
resented by the change of firing rate of the rel-
evant neurons, but by the (near) constant
firing of certain neurons that represent the
motion. The figure is an analogy. It shows how
a static picture can suggest motion.
© 2003 Nature Publishing Group
nature neuroscience • volume 6 no 2 • february 2003 123
8. Attention and binding
Attention can usefully be divided into two
forms: either rapid, saliency-driven and
bottom-up or slower, volitionally con-
trolled and top-down. Each form of atten-
tion can also be more diffuse or more
focused. Attention probably acts by bias-
ing the competition among rival coali-
tions, especially during their formation24.
Bottom-up attention may often start from
certain layer 5 neurons that project to parts
of the thalamus and the superior collicu-
lus. Top-down attention from the front of
the brain may go by somewhat diffuse
back-projections to apical dendrites in lay-
ers I, II and III, and perhaps also via the
intralaminar nuclei of the thalamus
(because these have inputs from the front
of the brain). Although such projections
are widespread, it does not follow that they
are nonspecific. To attend to ‘red’ involves
specific connections to many places in cor-
tex. An attractive hypothesis is that the
thalamus is largely the organ of attention.
The reticular nucleus of the thalamus may
help select among attentional signals on a
broad scale. Although attention can pro-
duce consciousness of a particular object
or event by biasing competition among
coalitions, activities associated with non-
attended objects are quite transient, giv-
ing rise to fleeting consciousness (such as
the proto-objects suggested in ref. 42).
What is binding? (For reviews that
address the binding problem, see Neuron
24, 1999.) This is the term used for the
process that brings together rather differ-
ent aspects of an object/event, such as its
shape, color, movement and so on. Bind-
ing can be of several types11. If it has been
laid down epigenetically, or has been learnt
by experience, it is already embodied in
one or more essential nodes so that no spe-
cial binding mechanism is needed. If the
binding required is (relatively) novel, then
in some way the activities of separate essen-
tial nodes must be made to act together.
Recent psychophysics suggests that
parallel versus serial’ search and ‘pre-
9. Styles of firing
Synchronized firing (including various
oscillations) may increase the effectiveness
of a neuron, while not necessarily altering
its average firing rate43. The extent and
significance of synchronized firing in cor-
tex remains controversial44. Computations
show45 that this effectiveness is likely to
depend on how the correlated input influ-
ences the excitatory and inhibitory neu-
rons in the recipient region to which the
synchronized neurons project.
We no longer think11 that synchro-
nized firing, such as the so-called 40 Hz
oscillations, is a sufficient condition for
the NCC. One likely purpose of synchro-
nized firing is to assist a nascent coalition
in its competition with other (nascent)
coalitions. If the visual input is simple,
such as a single bar in an otherwise empty
field, there might not be any significant
competition, and synchronized firing may
not occur. Similarly, such firing may not
be needed once a successful coalition has
reached consciousness, when it may be
able to maintain itself without the assis-
tance of synchrony, at least for a time46.
An analogy: after obtaining tenure, you
can relax a little.
At any essential node, the earliest spike
to arrive may sometimes have the advan-
tage over spikes arriving shortly there-
after47. In other words, the exact timing
of a spike may influence the competition.
attentive versus attentive’ processing
describe two independent dimensions
rather than variations along a single
dimension (Reddy, L., VanRullen, R. &
C.K., Vision Sci. Soc. 2nd Annu. Mtg.,abstr.
443, 2002). Their results can all be
expressed in terms of the relevant neural
networks. Several objects/events can be
handled simultaneously—more than one
object/event can be attended to at the
same time—if there is no significant over-
lap in any cortical neural network. That
is, if two or more objects/events do not
have any very active essential nodes in
common, they can be consciously per-
ceived. Under such conditions, several
largely separate (sensory) coalitions may
exist. If there is necessarily such an over-
lap, then (top-down) attention is needed
to select one of them by biasing the com-
petition among them.
This approach largely solves the clas-
sical binding problem, which was main-
ly concerned with how two different
objects/events could be ‘bound’ simul-
taneously. On this view, the ‘binding’ of
the features of a single object/event is
simply the membership in a particular
coalition. There is no single cortical area
where it all comes together. The effects
of that coalition are widely distributed
over both the back and the front of the
brain. Thus, effectively, they bind by
interacting in a diffuse manner.
Fig. 2. The dendritic arborization of the dif-
ferent types of neurons in the inferior tempo-
ral gyrus of the macaque monkey that project
to the prefrontal cortex near the principal sul-
cus (top, shaded gray). The neurons were
recovered from four slices (A–D) as indicated.
There are other types of neurons in this area
that project to other places. Note that only
one type of cell has apical dendrites that reach
to layer 1. Drawing from de Lima, A.D., Voigt,
T. and Morrison, J.H. (1990)39, reprinted with
permission of Wiley-Liss, Inc., a subsidiary of
John Wiley & Sons, Inc.
© 2003 Nature Publishing Group
124 nature neuroscience • volume 6 no 2 • february 2003
10. Penumbra and meaning
Consider a small set of neurons that fires
to, say, some aspect of a face. The experi-
menter can discover what visual features
interest such a set of neurons, but how
does the brain know what that firing rep-
resents? This is the problem of ‘meaning’
in its broadest sense.
The NCC at any one time will only
directly involve a fraction of all pyrami-
dal cells, but this firing will influence
many neurons that are not part of the
NCC. These we call the ‘penumbra’. The
penumbra consists of both synaptic effects
and also firing rates. The penumbra is not
the result of just the sum of the effects of
each essential node separately, but the
effects of that NCC as a whole. This
penumbra includes past associations of
NCC neurons, the expected consequences
of the NCC, movements (or at least pos-
sible plans for movement) associated with
NCC neurons, and so on. For example, a
hammer represented in the NCC is likely
to influence plans for hammering.
The penumbra, by definition, is not
itself conscious, although part of it may
become part of the NCC as the NCC
shifts. Some of the penumbra neurons
may project back to parts of the NCC, or
its support, and thus help to support the
NCC. The penumbra neurons may be the
site of unconscious priming48.
Related ideas
In the last 15–20 years, there has been an
immense flood of books and papers
about consciousness. For an extensive
bibliography, see T. Metzinger’s home-
metzinger. Many people have said that
consciousness is ‘global’ or has a ‘unity’
(whatever that is), but have provided few
details about such unity49. For many
years, Baars50 has argued that conscious-
ness must be widely distributed.
We are not receptive to physicists try-
ing to apply exotic physics to the brain,
about which they seem to know very lit-
tle, and even less about consciousness.
Dennett51,52 has written at length
about his ideas of “multiple drafts”, often
using elaborate analogies, but he seems not
to believe in the existence of consciousness
in the same way as we do. Dennett does
consider a limited number of psychologi-
cal experiments, but neurons, he tells us,
“are not my department” (pers. comm.).
Grossberg53 has written for many years
about his “adaptive resonance theory”
(ART). This he has developed using very
simple neural models. ART involves inter-
actions between the forward and the back
The idea of snapshots is a guess at the
dynamic properties of the parts of a suc-
cessful coalition, as coalitions are not sta-
tic, but constantly changing. The
penumbra, on the other hand, is all the
neural activity produced by the current
NCC, yet not strictly part of it.
We also speculate that the actual NCC
may be expressed by only a small set of
neurons, in particular those that project
from the back of cortex to those parts of
the front of cortex that are not purely
motor and that receive feedback from
there. However, there is much neural
activity leading up to and supporting the
NCC, so it is important to study this as
well as the NCC proper. Moreover, dis-
covering the temporal sequence of such
activities (A precedes B) will help us to
move from correlation to causation.
The explanation here of this inter-
locking set of ideas is necessarily abbrevi-
ated. We have outlined some of this in
more detail elsewhere10. A more extend-
ed account, together with descriptions of
the key experimental data, will be pub-
lished in a forthcoming book by C.K.56.
The above framework is a guide to con-
structing more detailed hypotheses so that
they can be tested against already-existing
experimental evidence and, above all, to
suggest new experiments. The aim is to
couch all such explanations in terms of the
behavior of identified neurons and in the
dynamics of very large neural assemblies.
Future experiments
These fall under several headings. Much
further experimental work on small
groups of neurons is required for cases in
which the percept differs significantly from
the sensory input, such as in binocular
rivalry57 and in the many visual illusions58.
Knowledge of the detailed neu-
roanatomy of the cerebral cortex needs to
be greatly expanded, in particular to char-
acterize the many different types of pyra-
midal cells in any particular cortical area.
What do they look like, where do they pro-
ject and, eventually, does each have a set
of characteristic genetic markers? Are there
types of pyramidal cells that do not occur
in all cortical areas? When recording spik-
ing activity from a neuron, it would be
very desirable to know what type it is, and
to where this particular cell projects.
The anatomical methods to character-
ize types of neurons for the macaque
monkey have been available for some
years39 (Fig. 2). In the last two decades,
very little work has been carried out on
cell types and their connectivity, mainly
for lack of funds, since such work is not
pathways, but there is little reference to
There are two fairly recent books
expressing ideas that overlap considerably
with ours. The first of these is by Edelman
and Tononi23. Their “dynamic core” is
very similar to our coalitions. They also
divide consciousness into primary con-
sciousness (which is what we are mainly
concerned with) and higher-order con-
sciousness (which we have, for the
moment, put on one side). They state
strongly, however, that they don’t think
there is a special subset of neurons that
alone expresses the NCC.
The second book54 is by Bachmann,
who for many years has used the term
“microgenesis” to mean what is happen-
ing in the 100–200 ms leading up to con-
sciousness, which is our own main concern
as well. He considers carefully the relevant
psychological phenomena, such as the dif-
ferent types of masking, but has fewer ideas
about the detailed behavior of neurons.
A framework somewhat similar to
ours has recently been described by
Dehaene and Naccache55, though they do
not elaborate on the snapshot hypothesis
or on the neural basis of essential nodes.
General remarks
Almost all of the above ideas have been
mentioned previously, either by us or by
others. We have given references to some
of these. We believe that the framework
we have proposed knits all these ideas
together, so that for the first time we have
a coherent scheme for the NCC in philo-
sophical, psychological and neural terms.
What ties all these suggestions togeth-
er is the idea of competing coalitions. The
illusion of a homunculus inside the head
looking at the sensory activities of the
brain suggests that the coalition(s) at the
back are in some way distinct from the
coalition(s) at the front. The two types of
coalitions interact extensively, but not
exactly reciprocally.
Zombie modes show that not all
motor outputs from the cortex are carried
out consciously. Consciousness depends
on certain coalitions that rest on the prop-
erties of very elaborate neural networks.
We consider attention to consist of mech-
anisms that bias the competition among
these nascent coalitions.
We suggest that each node in these
networks has a characteristic behavior.
We speculate that the smallest group of
neurons to be worth considering as a
node is a cortical column, with its own
characteristic behavior (its receptive and
projective fields).
© 2003 Nature Publishing Group
nature neuroscience • volume 6 no 2 • february 2003 125
‘hypothesis-driven. However, because
structure is often a clue to function,
detailed neuroanatomy is essential back-
ground knowledge (of the type the
human genome project provides for mol-
ecular biology).
On occasion, multi-unit electrodes are
chronically implanted into alert patients
(for example, to localize seizure onset
areas in epileptic patients). With their
consent, this can provide sparse but criti-
cal data about the behavior of neurons
during conscious perception or imagery59.
It would be very valuable if cortical tissue
could be stimulated appropriately with
such electrodes to generate specific per-
cepts, thoughts or actions60.
To study the dynamics of coalitions
requires simultaneous recordings on a fast
time scale, from small groups of neurons
in many places in the brain. This might
be done on a primate with a relatively
smooth cortex, such as the owl monkey.
It would require recording simultaneous-
ly from probably a thousand or more elec-
trodes, spaced about 1 mm apart, each
capable of picking up spikes from single
cells as well as the local field potential.
The immense amount of data this
would produce could be displayed visually
on a two-dimensional map of the cortical
surface, either speeded up or slowed down,
so that the eye could grasp the nature of
the traveling net-waves as a preliminary to
a more detailed study of them. The tech-
nical difficulties in recording from so many
neurons at once in an alert animal are for-
midable but not insuperable. To develop
the method, one could start with a smaller
number of electrodes, more widely spaced,
and work on one side of the brain of an
animal with the corpus callosum cut.
Some major omissions from this discus-
sion are (i) a more detailed consideration
of the role of the thalamus (and especial-
ly of the intralaminar nuclei), (ii) the
actions of the basal ganglia and of the dif-
fuse inputs from the brain stem and (iii)
a more detailed scheme for the overall
organization of the front of the cortex and
for motor outputs. For example, is there
some sort of hierarchy in the front? Are
there separate streams of information, as
there are in the visual system? Do the neu-
rons in the front of cortex show columnar
behavior and if so, what for?
On the other hand, our concentration
on the NCC, and the postponement of the
so-called hard problem of qualia, is part
of our strategic approach to the overall
problem of consciousness.
homunculus. in The Neural Correlates of
Consciousness (ed. Metzinger, T.) 103–110
(MIT Press, Cambridge, Massachusetts,
18. Jackendoff, R. Consciousness and the
Computational Mind (MIT Press, Cambridge,
Massachusetts, 1987).
19. Jackendoff, R. How language helps us think.
Pragmat. Cogn. 4,1–34 (1996).
20. Crick, F. & Koch, C. The unconscious
homunculus. Neuro-psychoanalysis 2,3–11
and subsequent pages for multi-authored
discussion (2000).
21. Rossetti, Y. Implicit short-lived motor
representations of space in brain damaged and
healthy subjects. Conscious. Cogn. 7,520–558
22. Hebb, D. The Organization of Behavior: a
Neuropsychological Theory (John Wiley, New
Yor k, 1949).
23. Edelman, G.M. & Tononi, G. A Universe of
Consciousness (Basic Books, New York, 2000).
24. Desimone, R. & Duncan, J. Neural
mechanisms of selective visual attention.
Annu. Rev. Neurosci. 18,193–222 (1995).
25. Wegner, D. The Illusion of Conscious Will (MIT
Press, Cambridge, Massachusetts, 2002).
26. Newsome, W.T. & Pare, E.B. A selective
impairment of motion perception following
lesions of the middle temporal visual area
(MT). J. Neurosci. 8,2201–2211 (1988).
27. Zeki, S.M. Parallel processing, asynchronous
perception, and a distributed system of
consciousness in vision. Neuroscientist 4,
365–372 (1998).
28. Zeki, S. & Bar tels, A. Toward a theory of visual
consciousness. Conscious. Cogn. 8,225–259
29. Mountcastle, V.B. Percep tual Neuroscience
(Harvard Univ. Press, Cambridge,
Massachusetts, 1998).
30. Biederman, I. Perceiving real-world scenes.
Science 177,77–80 (1972).
31. Wolfe, J.M. & Bennett, S.C. Preattentive object
files: shapeless bundles of basic features.
Vision Res. 37,25–43 (1997).
32. Hochstein, S. & Ahissar, M. View from the
top: hierarchies and reverse hierarchies in the
visual system. Neuron 36, 791–804 (2002).
33. Sherman, S.M. & Guillery, R. Exploring the
Thalamus (Academic Press, San Diego, 2001).
34. Zihl, J., Von Cramon, D. & Mai, N. Selective
disturbance of movement vision after bilateral
brain damage. Brain 106,313–340 (1983).
35. Hess, R.H., Baker, C.L. Jr. & Zihl, J. The
motion-blind’ patient: low-level spatial and
temporal filters. J. Neurosci. 9,1628–1640
36. Simpson, W.A. Temporal summation of visual
motion. Vision Res. 34,2547–2559 (1994).
37. Varela, F.J., Toro, A., John, E.R. & Schwartz,
E.L. Perceptual framing and cortical alpha
rhythm. Neuropsychologia 19,675–686 (1981).
38. Stroud, J.M. The fine structure of
psychological time. in Information Theory in
Psycholog y (ed. Quasten, H.) 174–207 (Free
Press, Glencoe, Illinois, 1955).
39. de Lima, A.D., Voigt, T. & Morrison, J.H.
Morphology of the cells within the inferior
temporal gyrus that project to the prefrontal
cortex in the macaque monkey. J. Comp.
Neurol. 296,159–172 (1990).
We thank P.S. Churchland, D. Eagleman,
G. Kreiman, N. Logothetis, G. Mitchison, T. Poggio,
V. Ramachandran, A. Revonsuo and J. Reynolds for
thoughtful comments, O. Crick for the drawing and
the J.W. Kieckhefer Foundation, the W.M. Keck
Foundation Fund for Discovery in Basic Medical
Research at Caltech, the National Institutes of
Health, the National Institute of Mental Health
and the National Science Foundation for
financial support.
1. Chalmers, D.J. The Conscious Mind: in Search
of a Fundamental Theory (Oxford Univ. Press,
New York, 1995).
2. Shear, J. Explaining Consciousness: the Hard
Problem (MIT Press, Cambridge,
Massachusetts, 1997).
3. Crick, F.C. & Koch, C. Consciousness and
neuroscience. Cereb. Cortex 8,97–107 (1998).
4. Crick, F.C. & Koch, C. Why neuroscience may
be able to explain consciousness. Sci. Am. 273,
84–85 (1995).
5. Allman, J., Miezin, F. & McGuinness, E.
Stimulus specific responses from beyond the
classical receptive field: neurophysiological
mechanisms for local–global comparisons in
visual neurons. Annu. Rev. Neurosci. 8,
407–430 (1985).
6. Lehky, S.R. & Sejnowski, T.J. Network model
of shape-from-shading: neural function arises
from both receptive and projective fields.
Nature 333,452–454 (1988).
7. Graziano, M.S., Taylor, C.S. & Moore, T.
Complex movements evoked by
microstimulation of precentral cortex. Neuron
34,841–851 (2002).
8. Felleman, D.J. & Van Essen, D.C. Distributed
hierarchical processing in the primate cerebral
cortex. Cereb. Cortex 1,1–47 (1991).
9. Poggio, T., Torre, V. & Koch, C.
Computational vision and regularization
theory. Nature 317,314–319 (1985).
10. Crick, F.C. & Koch, C. What are the neural
correlates of consciousness? in Problems in
Systems Neuroscience (eds. van Hemmen, L. &
Sejnowski, T.J.) (Oxford Univ. Press, New
Yor k, 2003).
11. Crick, F.C. & Koch, C. Towards a
neurobiological theory of consciousness. Sem.
Neuro sci. 2,263–275 (1990).
12. Crick, F. & Koch, C. Are we aware of neural
activity in primary visual cortex? Nat ure 375,
121–123 (1995).
13. Crick, F. & Koch, C. Constraints on cortical
and thalamic projections: the no-strong-loops
hypothesis. Natu re 391,245–250 (1998).
14. Milner, D.A. & Goodale, M.A. The Visual
Brain in Action (Oxford Univ. Press, Oxford,
UK, 1995).
15. Koch, C. & Crick, F.C. On the zombie within.
Nature 411,893 (2001).
16. Attneave, F. In defense of homunculi. in
Sensory Communication (ed. Rosenblith,
W.A.) 777–782 (MIT Press and John Wiley,
New York, 1961).
17. Crick, F.C. & Koch, C. The unconscious
© 2003 Nature Publishing Group
126 nature neuroscience • volume 6 no 2 • february 2003
40. Edelman, G.M. The Remembered Present: a
Biological Theory of Consciousness (Basic
Books, New York, 1989).
41. Purves, D., Paydartar, J.A. & Andrews, T.J. The
wagon wheel illusion in movies and reality.
Proc. Natl. Acad. Sci. USA 93,3693–3697
42. Rensink, R.A. Seeing, sensing and
scrutinizing. Vision Res. 40,1469–1487
43. Singer, W. & Gray, C.M. Visual feature
integration and the temporal correlation
hypothesis. Annu. Rev. Neurosci. 18,555–586
44. Shadlen, M.N. & Movshon, J.A. Synchrony
unbound: a critical evaluation of the temporal
binding hypothesis. Neuron 24,67–77,
111–125 (1999).
45. Salinas, E. & Sejnowski, T.J. Correlated
neuronal activity and the flow of neural
information. Nat. Rev. Neurosci. 2,539–550
46. Revonsuo, A., Wilenius-Emet, M., Kuusela, J.
54. Bachmann, T. Microgenetic Approach to the
Conscious Mind (John Benjamins, Amsterdam
and Philadelphia, 2000).
55. Dehaene, S. & Naccache, L. Towards a
cognitive neuroscience of consciousness: basic
evidence and a workspace framework.
Cognition 79,1–37 (2001).
56. Koch, C. The Quest for Consciousness: a
Neurobiological Approach (Roberts and
Company Publishers, California, in press).
57. Leopold, D.A. & Logothetis, N.K. Multistable
phenomena: changing views in perception.
Tre nds Cogn. Sci. 3,254–264 (1999).
58. Eagleman, D.M. Visual illusions and
neurobiology. Nat. Rev. Neurosci. 2,920–926
59. Kreiman, G., Fried, I. & Koch, C. Single-
neuron correlates of subjective vision in the
human medial temporal lobe. Proc. Natl.
Acad. Sci. USA 99,8378–8383 (2002).
60. Fried, I., Wilson, C.L., MacDonald, K.A. &
Behnke, E.J. Electric current stimulates
laughter. Nature 391,650 (1998).
& Lehto, M. The neural generation of a
unified illusion in human vision. Neuroreport
8,3867–3870 (1997).
47. Van Rullen, R. & Thorpe, S.J. The time course
of visual processing: from early perception to
decision-making. J. Cogn. Neurosci. 13,
454–461 (2001).
48. Schacter, D.L. Priming and multiple memory
systems: perceptual mechanisms of implicit
memory. J. Cogn. Neurosci. 4,255–256 (1992).
49. Bayne, T. & Chalmers, D.J. What is the unity of
consciousness? in The Unity of Consciousness:
Binding, Integration, Dissociation (ed.
Cleeremans, A.) (Oxford Univ. Press, Oxford,
UK, in press).
50. Baars, B.J. In the Theater of Consciousness
(Oxford Univ. Press, New York, 1997).
51. Dennett, D.C. Consciousness Explained (Little,
Brown & Co., Boston, Massachusetts, 1991).
52. Dennett, D. Are we explaining consciousness
yet? Cognition 79,221–237 (2001).
53. Grossberg, S. The attentive brain. Am. Sci. 83,
438–449 (1995).
© 2003 Nature Publishing Group
... Através do homem, criou-se um novo mundo objectivo, o mundo dos produtos da mente humana; um mundo de mitos, de contos de fadas e teorias científicas; de poesia, arte e música." 23 Esses produtos da mente são, não só informação de, mas também informação para. A consciência define-se numa visão tripartida: percepção dos estímulos exteriores ou interiores ao sistema (consciência do mundo), (re)acção do sistemas como capacidade de resposta (consciência da acção), reflexão do sistema como capacidade de reflectir tanto estímulos ou (re)acções (consciência da consciência). ...
... ... A filosofia da gramática e a do entendimento humano são aliadas mais próximas do que é habitualmente imaginado." 23 Por outro lado, a própria linguagem é construtora de realidades e está em permanente mutação. 24 Reid antecipa com génio a teoria dos actos de fala de Austin e de Searle. ...
Full-text available
Inclui estudos de Thomas Metzinger (Johannes Gutenberg-Universität Mainz), “Transparência Fenoménica e Autorreferência Cognitiva”, traduzido por Manuel Curado e Miguel Pais-Vieira; José António Alves (UCP), “Imaginação e Liberdade”; André Barata (UBI), “Da Experiência Mental sem Consciência ou o Problema Mente-Corpo para lá da Consciência de Acesso e da Consciência Fenomenal”; José Miguel Stadler Dias Costa (UCP), “Evolução da Mente Artística: Psicologia e Cognição”; Manuel Curado (Univ. Minho), “O Choque de Thomas Reid e a Origem do Problema Difícil da Consciência”; Alfredo Dinis, SJ (UCP), “Aspetos Científicos e Filosóficos do Estudo do Correlato Neural da Consciência”; António M. Fonseca (UCP), “A Conservação do Self no Decurso do Envelhecimento”; e Marta de Assunção Gonçalves (Univ. Évora/UCP), “O Cérebro Analfabeto: O Intrigante Caso do Não Reconhecimento de Desenhos”.
... Stages of Chetana There is an unfathomable length of canonical literature in the ancient scriptures like Tattvarthsutra, Sarvarthsiddhi , Pravachansaar (1st century) , Raajvaartik (7th century), etc. that explain the minute and subtle processing of knowing and feeling (to be aware in psychology) function of Chetana consciousness in 6 major stages (Raajvaartik pg. [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30]. ...
... In addition, the neurological perspective of consciousness by Krick and Koch states that, "the problem of consciousness can, in the long run, be solved only by explanations at the neural level" [21]. It suggested a rather boastful and desperate approach reducing the 'problem' of consciousness in spite of defining consciousness as a conceptual term 'phenomenology' by Husserl a century ago [22]. Krick and Koch further concluded their research regarding the oscillatory operations of visual awareness, which they considered as the most significant explanatory component of consciousness, by stating "A striking feature of our visual awareness (and of consciousness in general) is that it is very rich in information, even if much of it is retained for only a rather brief time. ...
Full-text available
The healing from any ailment entails the comprehensive management of the individual’s bio-psycho-socio-spiritual context. The effect of psychological domains on one’s physical health indicate the scope of study of consciousness. Scientific psychological research findings along with the prevailing ancient traditions (religion, philosophy, culture) of India, have given a green signal to the major role of consciousness in one’s well being. The present study aims to analyze the scope of psychosomatic healing with the perspective of Jain philosophy, which covers a vast multifold understanding of consciousness with the word ‘chetana’. Early canonical literature like Tattvarthsutra, Sarvarthasiddhi , Pravachansaar, Raajvaartik of Jain Philosophy have described 6 stages of Consciousness and 2 dimensions for its manifestation where cognition is a primary manifestation of the consciousness. It was seen that an absolute neutrality of all components of cognition and consciousness such as emotion, thought, comprehension, apprehension, feeling, perception, experience, etc. when measured and quantified as an absolute zero, reached the state of consciousness homeostasis; wherein a balanced level of consciousness leads to a balanced psychological state. The study presented a model of information processing of consciousness, as described in Jain philosophy for achieving consciousness homeostasis for psychotherapy, providing us ancient and novel principles for psychotherapeutics, and paving a way for an application-based consciousness theory that is transferable to a clinical setup in the dynamics of therapy.
... After all, conscious states emerge globally at a macroscale of the whole-brain network. This scale corresponds to a phenomenal (psyche) level, to which many volitional and cognitive systems contribute, thereby generating what is viewed as the neural correlates of consciousness (NCC) (Crick and Koch 2003). This psyche level is neither volitional nor cognitive but only representative of these both. ...
... To do it consistently, consider the concept of NCC in the framework of the stream. The NCC has been traditionally defined as the minimal neural substrate expressed by specific signatures that are necessary and sufficient for any conscious experience (Crick and Koch 2003). This is based on the assumption that a key function of consciousness is to produce the best current interpretation of the visual scene and to make this information available to the planning stages of the brain (Rees et al. 2002). ...
Full-text available
The brain integrates volition, cognition, and consciousness seamlessly over three hierarchical (scale-dependent) levels of neural activity for their emergence: a causal or ‘hard’ level, a computational (unconscious) or ‘soft’ level, and a phenomenal (conscious) or ‘psyche’ level respectively. The cognitive evolution theory (CET) is based on three general prerequisites: physicalism, dynamism, and emergentism, which entail five consequences about the nature of consciousness: discreteness, passivity, uniqueness, integrity, and graduation. CET starts from the assumption that brains should have primarily evolved as volitional subsystems of organisms, not as prediction machines. This emphasizes the dynamical nature of consciousness in terms of critical dynamics to account for metastability, avalanches, and self-organized criticality of brain processes, then coupling it with volition and cognition in a framework unified over the levels. Consciousness emerges near critical points, and unfolds as a discrete stream of momentary states, each volitionally driven from oldest subcortical arousal systems. The stream is the brain’s way of making a difference via predictive (Bayesian) processing. Its objective observables could be complexity measures reflecting levels of consciousness and its dynamical coherency to reveal how much knowledge (information gain) the brain acquires over the stream. CET also proposes a quantitative classification of both disorders of consciousness and mental disorders within that unified framework.
... Around the same time, the late Francis Crick was instrumental in bringing consciousness to the fore as a topic of legitimate (neuro)scientific investigation. He proposed that scientists should search for the "neural correlates of consciousness": identifying aspects of brain function that co-vary with changes in consciousness, as a starting point towards mapping the relationship between the two with increasing precision 13,14 . A more recent evolution of the scientific stance towards the neuroscientific study of consciousness is Anil Seth's "Real problem of consciousness" approach, which proposes to reframe the endeavour as aiming to "explain, predict, and control consciousness". ...
Different perturbations of the brain’s delicate functioning, ranging from transient pharmacological interventions to severe trauma, can result in altered states of consciousness. To illuminate how the neurobiology and organization of the human brain support consciousness, we need to identify changes in brain function that accompany alterations in conscious state. However, the brain is a paradigmatic example of a complex system, raising the question: which aspects of its complex functioning and architecture should be the focus of our investigation? Traditionally, the quest for the “neural correlates of consciousness” has been framed in terms of spatial localisation: which brain regions are most relevant for consciousness? Complementing this extensive body of work, in my thesis I consider three alternative ways of conceptualising brain function (quantified from functional MRI), and how it may support consciousness. First, I adopt a time-resolved perspective, decomposing brain activity into predominantly integrated or segregated patterns of dynamic functional connectivity. Building on my previous work in anaesthesia and disorders of consciousness, I show how the dynamic interplay of functional integration and segregation is reshaped by the classic serotonergic psychedelic, LSD. Second, I consider a frequency-resolved perspective, decomposing functional brain activity into patterns of structure-function coupling across scales: the harmonic modes of the human connectome. This “connectome harmonic decomposition” of brain activity reveals a generalisable neural signature of loss of consciousness, whether due to anaesthesia or brain injury. A mirror-reverse of this harmonic signature characterises the altered state induced by LSD or ketamine. Connectome harmonics provide a robust indicator of consciousness across datasets, correlating with physiological and subjective variables. On the theoretical side, neuroscientific theories postulate that consciousness depends on the integration of information by a “global workspace” of brain regions. However, these accounts treat “information” as a primitive, whereas the recent framework of information decomposition has shown that Shannon information is actually a composite of several more fundamental kinds of information, including synergistic information, which is available only when a set of sources are considered jointly, and redundant information, which is available from multiple individual sources. Demonstrating the importance of disentangling these different kinds of information, I develop a framework for information-resolved analysis of brain activity, based on information decomposition. Combining functional and diffusion MRI, PET, and transcriptomics, I show that higher cognitive systems in the brain leverage the efficiency of synergistic information, whereas redundant interactions are predominantly associated with modular, structurally-coupled sensorimotor systems. Finally, by explicitly taking into account these fundamental kinds of information, I formalise the “global workspace” architecture in information-theoretic terms, revealing that both anaesthesia and disorders of consciousness induce a breakdown of synergistic integration in the brain’s Default Mode Network. Conceptually, these results contribute to reconciling two prominent theories of consciousness, the Global Neuronal Workspace Theory and Integrated Information Theory. Overall, viewing the brain as a time-, frequency-, and information-resolved complex system offers fruitful new ways to understand the brain’s functional architecture, laying the foundations to map the rich landscape of human consciousness.
... Visual experience often serves as a basic example of conscious experience. Several scientists and philosophers have focused solely on the study of visual percepts, as a means of identifying the minimal set of neural events required to elicit a conscious mental experience [13]. ...
Full-text available
Ahuna Mons is a 4 km particular geologic feature on the surface of Ceres, of possibly cryovolcanic origin. The special characteristics of Ahuna Mons are also interesting in regard of its surrounding area, especially for the big crater beside it. This crater possesses similarities with Ahuna Mons including diameter, age, morphology, etc. Under the cognitive psychology perspective and using current computer vision models, we analyzed these two features on Ceres for comparison and pattern-recognition similarities. Speeded up robust features (SURF), oriented features from accelerated segment test (FAST), rotated binary robust independent elementary features (BRIEF), Canny edge detector, and scale invariant feature transform (SIFT) algorithms were employed as feature-detection algorithms, avoiding human cognitive bias. The 3D analysis of images of both features’ (Ahuna Mons and Crater B) characteristics is discussed. Results showed positive results for these algorithms about the similarities of both features. Canny edge resulted as the most efficient algorithm. The 3D objects of Ahuna Mons and Crater B showed good-fitting results. Discussion is provided about the results of this computer-vision-techniques experiment for Ahuna Mons. Results showed the potential for the computer vision models in combination with 3D imaging to be free of bias and to detect potential geoengineered formations in the future. This study also brings forward the potential problem of both human and cognitive bias in artificial-intelligence-based models and the risks for the task of searching for technosignatures.
... To visualize the temporal distribution more directly, individual maps (Figure S2A) were combined into a 3,702 biopsy-site-level temporo-spatial atlas, where each brain region is color coded by the branch with the smallest p value (u peak) (see Figure 2 for representative sections and Data S3 for a brain-wide atlas). Ranked u peaks from early primate ancestors ( Figure 2; branch 1; medulla, hippocampal formation, and claustrum) may reflect the evolution of basic cognitive traits, such as autonomic regulation, spatial memory, attention, and consciousness (claustrum; Smith et al., 2019;Crick and Koch, 2003; see Table S11). In contrast, ranked u peaks from hominin branches expanded to higher, cortical areas (branches 6, 7, and AMH; Figure 2). ...
Full-text available
The brains and minds of our human ancestors remain inaccessible for experimental exploration. Therefore, we reconstructed human cognitive evolution by projecting nonsynonymous/synonymous rate ratios (ω values) in mammalian phylogeny onto the anatomically modern human (AMH) brain. This atlas retraces human neurogenetic selection and allows imputation of ancestral evolution in task-related functional networks (FNs). Adaptive evolution (high ω values) is associated with excitatory neurons and synaptic function. It shifted from FNs for motor control in anthropoid ancestry (60–41 mya) to attention in ancient hominoids (26–19 mya) and hominids (19–7.4 mya). Selection in FNs for language emerged with an early hominin ancestor (7.4–1.7 mya) and was later accompanied by adaptive evolution in FNs for strategic thinking during recent (0.8 mya–present) speciation of AMHs. This pattern mirrors increasingly complex cognitive demands and suggests that co-selection for language alongside strategic thinking may have separated AMHs from their archaic Denisovan and Neanderthal relatives.
... The concept of an "organ of attention" is not new: many scientists have already started investigating the neural and brain structures constituting it (Mesulam, 1990;Posner and Petersen, 1990;Crick, 1994;Crick and Koch, 2003). However, the search for such an organ is not fully uncontroversial. ...
Full-text available
What distinguishes conscious information processing from other kinds of information processing is its phenomenal aspect (PAC), the-what-it-is-like for an agent to experience something. The PAC supplies the agent with a sense of self, and informs the agent on how its self is affected by the agent’s own operations. The PAC originates from the activity that attention performs to detect the state of what I define “the self” (S). S is centered and develops on a hierarchy of innate and acquired values, and is primarily expressed via the central and peripheral nervous systems; it maps the agent’s body and cognitive capacities, and its interactions with the environment. The detection of the state of S by attention modulates the energy level of the organ of attention (OA), i.e., the neural substrate that underpins attention. This modulation generates the PAC. The PAC can be qualified according to five dimensions: qualitative, quantitative, hedonic, temporal and spatial. Each dimension can be traced back to a specific feature of the modulation of the energy level of the OA.
In Holoplexity, consciousness is hypothesized as predating the universe and as ultimately comprising all matter and energy. It comprises the very architecture of reality, including spatial dimensions (as we perceive them), and is even the causal factor of time itself. This theory then goes on to argue that consciousness is hidden from us, is timeless (but still generates time), and is the source from which all things flow. Humans are only able to appreciate and apprehend the aftermath of this interaction. Consciousness is then believed to exist in all things as manifested in both matter and electromagnetism, as well as non-spatial, non-temporal, phenomenal existence itself. Holoplexity seeks to offer an explanation for how information becomes human experience. From the advent of time to the reading of these words, the Holoplexity Theory of Consciousness makes a coherent explanation for it all.
Several cortical and subcortical brain areas have been reported to be sensitive to the emotional content of subliminal stimuli. However, the timing of these activations remains unclear. Our scope was to detect the earliest cortical traces of emotional unconscious processing of visual stimuli by recording event-related potentials (ERPs) from 43 participants. Subliminal spiders (emotional) and wheels (neutral), sharing similar low-level visual parameters, were presented at two different locations (fixation and periphery). The differential (peak-to-peak) amplitude from CP1 (77 ms from stimulus onset) to C2 (100 ms), two early visual ERP components originated in V1/V2 according to source localization analyses, was analyzed via Bayesian and traditional frequentist analyses. Spiders elicited greater CP1–C2 amplitudes than wheels when presented at fixation. This fast effect of subliminal stimulation—not reported previously to the best of our knowledge—has implications in several debates: 1) The amygdala cannot be mediating these effects, 2) latency of other evaluative structures recently proposed, such as the visual thalamus, is compatible with these results, 3) the absence of peripheral stimuli effects points to a relevant role of the parvocellular visual system in unconscious processing.
Part One * Consciousness and the Scientific Observer * Proposals and Disclaimers Part Two * Neural Darwinism * Reentrant Signaling * Perceptual Experience and Consciousness Part Three * Memory as Recategorization * Time and Space: Cortical Appendages and Organs of Succession * Concepts and Presyntax Part Four * A Model of Primary Consciousness * Language * Higher-Order Consciousness * The Conscious and the Unconscious * Diseases of Consciousness Part Five * Physics, Evolution, and Consciousness: A Summary * Philosophical Issues: Qualified Realism * Epilogue
Visual awareness is a favorable form of consciousness to study neurobiologically. We propose that it takes two forms: a very fast form, linked to iconic memory, that may be difficult to study; and a somewhat slower one involving visual attention and short-term memory. In the slower form an attentional mechanism transiently binds together all those neurons whose activity relates to the relevant features of a single visual object. We suggest this is done by generating coherent semi-synchronous oscillations, probably in the 40-70 Hz range. These oscillations then activate a transient short-term (working) memory. We outfit several lines of experimental work that might advance the understanding of the neural mechanisms involved. The neural basis of very short-term memory especially needs more experimental study.
In recent years, many new cortical areas have been identified in the macaque monkey. The number of identified connections between areas has increased even more dramatically. We report here on (1) a summary of the layout of cortical areas associated with vision and with other modalities, (2) a computerized database for storing and representing large amounts of information on connectivity patterns, and (3) the application of these data to the analysis of hierarchical organization of the cerebral cortex. Our analysis concentrates on the visual system, which includes 25 neocortical areas that are predominantly or exclusively visual in function, plus an additional 7 areas that we regard as visual-association areas on the basis of their extensive visual inputs. A total of 305 connections among these 32 visual and visual-association areas have been reported. This represents 31% of the possible number of pathways it each area were connected with all others. The actual degree of connectivity is likely to be closer to 40%. The great majority of pathways involve reciprocal connections between areas. There are also extensive connections with cortical areas outside the visual system proper, including the somatosensory cortex, as well as neocortical, transitional, and archicortical regions in the temporal and frontal lobes. In the somatosensory/motor system, there are 62 identified pathways linking 13 cortical areas, suggesting an overall connectivity of about 40%. Based on the laminar patterns of connections between areas, we propose a hierarchy of visual areas and of somato sensory/motor areas that is more comprehensive than those suggested in other recent studies. The current version of the visual hierarchy includes 10 levels of cortical processing. Altogether, it contains 14 levels if one includes the retina and lateral geniculate nucleus at the bottom as well as the entorhinal cortex and hippocampus at the top. Within this hierarchy, there are multiple, intertwined processing streams, which, at a low level, are related to the compartmental organization of areas V1 and V2 and, at a high level, are related to the distinction between processing centers in the temporal and parietal lobes. However, there are some pathways and relationships (about 10% of the total) whose descriptions do not fit cleanly into this hierarchical scheme for one reason or another. In most instances, though, it is unclear whether these represent genuine exceptions to a strict hierarchy rather than inaccuracies or uncertainties in the reported assignment.
Descriptions of physical properties of visible surfaces, such as their distance and the presence of edges, must be recovered from the primary image data. Computational vision aims to understand how such descriptions can be obtained from inherently ambiguous and noisy data. A recent development in this field sees early vision as a set of ill-posed problems, which can be solved by the use of regularization methods. These lead to algorithms and parallel analog circuits that can solve ‘ill-posed problems’ and which are suggestive of neural equivalents in the brain.