June 17, 2005 8:59 WSPC/179-JIN 00076

Journal of Integrative Neuroscience, Vol. 4, No. 2 (2005) 183205

c Imperial College Press

Research Report

EMERGING CONCEPTS OF BRAIN FUNCTION

PAUL BACH-Y-RITA
Departments of Orthopedics and Rehabilitation Medicine
and of Biomedical Engineering, 1300 University Avenue, Room 2756
University of Wisconsin, Madison, WI, USA
pbachyri@wisc.edu

Received 14 September 2004

Accepted 24 December 2004

For over 40 years, since I rst obtained evidence for nonsynaptic diusion neurotransmis-
sion (most scientists call it Volume Transmission), I have been convinced that we scientists
were ignoring organizational dynamics other than the mechanistic synaptic organization
of the brain. For many years it was an uneasy feeling, since I was aware there are so many
avenues to explore in brain function. I have wondered how much we scientists have ignored,
in our quest to understand how the brain really works, due to our eorts to be scientic.
In addition to the diculty of understanding how the brain functions, how could we even
begin to explore the human experience? In this paper I will rst discuss some emerging
concepts of brain function. I will then comment on the development of concepts that have
been a part of my own research experience.

Keywords: Volume transmission; nonsynaptic; brain function; brain plasticity; functional

reorganization.

1. Introduction
The connectionist-localizationist theory of brain function, based on the neuron doc-
trine, was the product of many brilliant studies by Broca [37], Ramon y Cajal [38],
Sherrington [85], and other scientists of the last half of the 19th and the rst third of
the 20th century. Anatomical, physiologic, neurochemical and clinical data collected
during and shortly after that period t comfortably within that concept of how the
brain works.
However, that model was developed and established before the era of molec-
ular biology, before extrasynaptic diusion neurotransmission, before G-proteins
and ionic and metabotropic membrane signaling systems, before co-transmitters,
before glial involvement in neurotransmission, before the reappearance of adult
brain plasticity concepts, but after a century of domination of the neurosciences
by localizationism. As each of these and other processes were demonstrated over the

183
June 17, 2005 8:59 WSPC/179-JIN 00076

184 Bach-y-Rita

last third of a century, their roles have been interpreted within the existing concep-
tual framework. The t has not always been comfortable [10, 14, 16, 84], but since
there really was no alternative viable theory, the old paradigm has endured.

2. Persistence of the Connectionist Model

A principal aspect of this conceptual framework has been the domination of the
synaptic model of information transmission. However, synaptic transmission may not
be quantitatively the principal means of neurotransmission in the brain; Herkenham
[61] considers that receptor-transmitter release site mismatches (in comparison to
synapses, at which they are in close opposition) are the rule rather than the excep-
tion. The exclusivity of the synapse as a means of transmitting information was
thrown into doubt when results obtained from intra- and extracellular microelec-
trode studies of polysensory brain stem neurons could not be t into the prevailing
connectionist theory of brain function [5].
Freeman considers that the common view of the brain organization, as consist-
ing of neuronal networks, has been inuenced by the usual Golgi study illustrations
of brain structure. However, the Golgi method stains only about 1.5% of the cells,
thus giving . . . an erroneous impression of a discrete network of neurons instead
of a densely packed continuous tissue of axons, dendrites and synapses forming a
neuropil . . . Neuronal populations do not exist merely because of large numbers
of neurons driven by common input in parallel. They arise by virtue of feedback
interactions among cortical neurons that yield cooperative activity [50]. Similarly,
Gilbert, Hirsch and Wiesel [56, 57] consider that the evidence from earlier Golgi stud-
ies lead to the idea that cortical connections were primarily vertical, running across
the cortical layers and with relatively little lateral spread. Their studies showed that
horizontal processes of pyramidal cells (which allow integration over a large area of
the visual cortex) have sparse connections between individual cells; they operate
by mass action. Rather than thinking of receptive elds as being restricted in
their extent, with the process of integration of the components of an image occur-
ring at a much later stage along the visual pathway, they have shown that the
integrative process is a progressive one, beginning in the primary visual cortex (or
perhaps even earlier) and building up in a cascade of diverging and converging
connections.
Freeman [49] suggests that perceptual as distinct from sensory information is
not expressed at the neuronal level, but rather in neuronal modules that consist
of assemblies of tens of thousands to hundreds of millions of neurons. Extensive
studies [49, 50, 51] primarily on the olfactory system as a model of brain function
and largely in regard to mechanisms producing electroencephalogram (EEG) signals,
has also led him to consider a role for Volume Transmission (VT) in interactions
between the cells in a neuron assembly and in synaptically mediated functions, such
as the action of norepinephrine in enabling synaptic changes during learning in
populations of neurons, by nonsynaptic diusive release from axons en passant. He
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 185

views synaptic function as . . . the light play of high notes in an organ recitation
against the sustained structure of base notes (VT), both carrying the activity like
two legs in walking [16]. Our studies [12, 16, 31] suggest that Freemans sustained
structure of the base notes may be sucient in those cases where sustained rather
than immediate action is required.
During the century of virtually total domination of the neurosciences by hard-
wired localizationist concepts that followed Brocas demonstration of functional
localization, some scientists presented evidence for plasticity, but their ndings were
generally ignored. Thus, in a well known study, Hubel and Wiesel [63] sewed the
eyelid of one eye in kittens, eliminating visual input to the retina and to the cor-
responding part of the visual cortex. When they removed the sutures, the kittens
showed no ability to see with the previously sutured eye and they remained perma-
nently blind (amblyopic) in that eye. However, Chow and Stewart [39] repeated their
study but added visual training (rehabilitation), and obtained morphological, phys-
iological and behavioral evidence for recovery, which demonstrated the plasticity of
the brain. Nevertheless, at that time the conceptual substance of the neurosciences
did not include plasticity, and their study was rarely cited. Our early brain plasticity
studies demonstrating the capacity of congenitally blind persons to be trained to
perceive visual information delivered to the skin via a modied television camera and
an array of vibro- or electrotactile stimulators [7, 11, 27] were also incongruent with
the prevailing localizationist concepts. Thus, the prevailing conceptual framework
aects the generation of knowledge at all stages, including the choice of experi-
ment, the experimental design, the experimental methods, and the interpretation
of the results. In addition, it aects the interest generated by the results, as evi-
denced by the minimal interest in the Chow and Stewart and other brain plasticity
publications.
The prevailing conceptual framework also aects the therapy for specic dis-
ease entities; the conclusions of Hubel and Wiesel [63] regarding the permanence of
amblyopia following lid suture were not likely to have inspired plasticity-based reha-
bilitation approaches to brain dysfunction due to developmentally-induced abnor-
malities. Similarly, although Franz and associates demonstrated recovery from brain
damage with appropriate rehabilitation in both animals and humans and published
their results in the Journal of the American Medical Association in 1915 (in which
they suggested that we should not speak of permanent paralysis in hemiplegic
patients, but rather of untreated paralysis), their study did not inuence the clini-
cal management of stroke patients [48]. Furthermore, cerebral ablation experiments
led to the long-dominant clinical concept of permanent loss of function following
stroke or traumatic brain damage; little recovery was expected within that con-
ceptual framework, and (consistent with Mertons [73] ideas on the self-fullling
prophecy) little recovery was obtained [16, 32, 33, 42].
Within that conceptual framework, even most practitioners in the eld of Neu-
rological Rehabilitation have sought functional adaptation to the disability resulting
from a brain lesion, rather than reorganization of function [16, 23]. Yet Luria [72]
June 17, 2005 8:59 WSPC/179-JIN 00076

186 Bach-y-Rita

noted that automatic and instinctive reorganizations after a lesion simply exclude
the dysfunctional part of the body, while with appropriate rehabilitation, including
conscious eort, function is reorganized to include the motorically dysfunctional
parts of the body, resulting (within certain limits) in functional recovery.
It appears that the neurosciences are slower to shift paradigms than other elds.
Horrobin [62] suggested that this may be due to the absence of an appreciation for
theory and hypotheses. He noted that the concept of the dependence of observa-
tion on theory and hypothesis is thoroughly understood in physics, chemistry and
related sciences; in these elds there is a balanced tension between hypothesis and
theory on one hand and experiment on the other. It is recognized that discovery in
science almost always begins either with a hypothesis or with an observation that
can be seen as anomalous against the background of a clearly dened theoretical
construct. Horrobin noted that the situation in the biomedical sciences could hardly
be more dierent. We treat ideas in the absence of observation as something that
must not be allowed to see the light of day. He further noted that we have no tra-
dition of analytic criticism of existing theoretical concepts, and as a result, general
beliefs persist within the biomedical community long after evidence is available to
quash them.
In the absence of broad-based support for theoretical approaches within the dis-
ciplines of the neurosciences, theories have often been imposed by the strength of
personality and the professional reputations of the proponents, instead of resulting
from open debate in the neuroscience literature. This in no way detracts from the
great scientic contributions of those scientists; rather, it reects the virtual dei-
cation of those persons by their followers. Thus, in addition to connectionism and
localizationism discussed above, other exclusive theories have been dicult to over-
turn. One of these is the Sherringtonian reex-based concept of cortical function.
When experimental evidence from sensory denervated primates demonstrated that
motor control could be organized and initiated by brain activity without sensory
input, Taub [87] concluded that the reex-based concept of cerebral function could
no longer be supported. However, he reports the anger and rejection with which he
was confronted due to this challenge to the hallowed Sherringtonian dogma, despite
the fact that there was no claim that reex-based and internally-initiated actions
are incompatible alternatives.
Theory that is congruent with modern neurobiology will inuence our under-
standing of mechanisms such as learning, plasticity, drug actions, recovery from
brain damage, and of mood, sleep, hunger and other mass sustained functions [12],
as well as inuencing developments in the elds of articial intelligence and neuronal
networks [4], which are connectivity-based at present. However, the intrinsic limi-
tations in the evolution of scientic thought will limit the speed at which radically
new (or old but previously ignored) experimentally supported concepts enter into
the conceptual substance of the neurosciences. We scientists are convinced that we
are scientic. Ultimately, fashion has been as dominant in science as it is in the
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 187

eld of ladies clothing; when the hemline goes up, everyone follows suit. Similarly,
when a concept such as synaptic communication to the exclusion of all else is rmly
established, there is pressure to conform to the conventional dogma.

3. Conceptual Development Based on My Research

3.1. Volume transmission or nonsynaptic
diusion neurotransmission
It is quite likely that the brain wastes neither energy nor space. Synaptic trans-
mission requires nerve bers and alternating cycles of activation-inactivation (nerve
spike), which allows the nerve spikes to occur in rapid sequence, providing ne
functional control. When such ne control is not needed (e.g., for mass sustained
functions such as hunger), it appears to be excessively costly to use the synaptic
mechanism. Thus, I suggested, based on these studies, a Law of the Conservation of
Space and Energy in the Brain [17].
An alternative process of communication between neurons, VT, can be modeled
by students in a university classroom [25], who can be equated to neurotransmit-
ter molecules in a vesicle. Upon release, they must go to specic other classrooms
spread throughout the campus (receptor sites). They ow out into the halls and the
grounds between buildings (extracellular uid), where they mix with other students
(neurotransmitter molecules) from other classrooms (vesicles) going to other target
classrooms. They walk (diuse) to their specic classrooms (receptors) which they
enter (bind). In contrast, synaptic communication would require each student to
travel to the next class on a motorized separate pathway, much like the enclosed
people-transporter bands at the De Gaulle airport in Paris.
VT includes the diusion through the extracellular uid of neurotransmitters
released at points that may be remote from the target cells, with the resulting
activation of extrasynaptic receptors (and possible intrasynaptic receptors reached
by diusion into the synaptic cleft). In contrast to the one-to-one, point-to-point
private intercellular synaptic communication in the brain, VT is a slow, one-to-
many, widespread intercellular communication [96]. VT may play many roles in
the brain in normal and abnormal activity, brain plasticity and drug actions. VT
mechanisms may predominate in some functions, probably including the mechanism
of action of psychotropic drugs [14]. Combinations of both synaptic communication
and diusion may be the general rule.
VT appears to be the principal means of neurocommunication in certain
mammalian brain regions such the greater limbic system [80] and may play an
important role in the organization and regulation of behavior by the core and
paracore regions of the brain [81]. It may also be an important mechanism in
the highest levels of the human brain. Music appreciation and intellectual func-
tions, for example, may not require the high-frequency (and energetically costly)
June 17, 2005 8:59 WSPC/179-JIN 00076

188 Bach-y-Rita

alternating cycles of activation-inactivation of synaptic transmission, and may be

largely replaced by VT [26]. It may play a role in recovery from brain damage,
which causes changes in neurotransmitter levels. Some neurotransmitter systems
are up-regulated, while others are down-regulated [95]. Drug therapy and rehabili-
tation may induce functional recovery by inuencing the aected neurotransmitter
systems.
VT plays a role in space and brain energetics, in drug actions, and in recovery
from brain damage. It may be the primary information transmission mechanism
in certain normal mass, sustained functions, such as sleep, vigilance, hunger, brain
tone and mood and certain responses to sensory stimuli. In addition, it appears to
be a signicant inuence in several abnormal functions, including mood disorders,
spinal shock, spasticity, shoulder-hand and autonomic dysreexia syndromes, and
drug addiction.
With the development of the techniques of molecular neurobiology, VT was ele-
gantly demonstrated, such as in the immuno-histochemical studies of Fuxe and his
collaborators [53]. Beaudet and Descarries [36] showed that the biogenic amines
released from nonsynaptic varicosities may act not only upon adjacent post-synaptic
surfaces, but also in tissue of more distant receptor elements. Neurochemical trans-
mission by extrasynaptic diusion had also been noted by Vizi, whose studies on
the nonsynaptic modulation of transmitter release [89, 90, 92] led him to propose [91]
that the essence of brain function (e.g., learning, thinking) lies not in variations of
neuronal circuitry (hardware) but rather within the chemical communication itself
which is partly wired (synaptic) and partly unwired (nonsynaptic) transmission. I
have reviewed these studies elsewhere [13, 16, 20, 22].

3.2. Brain mass and brain energetics

Assemblies of cells, or neuronal modules, have been postulated to form the basis of
many functions of the brain [47, 52, 59, 86]. The cells are separated by a signicant
volume of extracellular space [79], and are connected either by nerve bers (Hebb
[59] considered that each cell in the assembly is connected to all the other cells
in the module synaptically), by the action of neurotransmitters diusing through
the extracellular space, or by a combination of both. Cell-assembly connectivity
in these modules is likely to be varying combinations of synaptic and nonsynaptic
mechanisms, depending on the specic function.
In Hebbs view [59], waves of depolarization spread along an impressive network
of links, carrying information all over the assembly. Hebb did not consider how
much energy would be necessary to drive such a network. The estimate depends
on the nature of the links. Unmyelinated links require depolarization of the entire
link, while myelinated nerves require only depolarization of the short unmyelinated
segments at the nodes of Ranvier. As noted above, the cost of an action potential
(AP) has been calculated [3] to be about 1011 1012 adenosine triphosphate ATP
per cm2 of depolarized membrane, with a minimum cost of 106 ATP at a node of
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 189

Ranvier. A small (10,000 neuron) module would require approximately 1 joule per
liter of brain. Furthermore, the volume of the nerve bers necessary to fully connect
the brain synaptically would create a brain so large that it would not t in the
head [25].
The strictly hard-wired scheme is unfeasible, given available brain metabolic
resources. VT oers a less costly alternative to synaptic communication. Thus, rather
than purely synaptic process, it is more likely that a hybrid process, exhibiting
varying combinations of hard-wired (synaptic) and wireless (VT) connectivities,
would t into the available volume in the brain, with a signicant decrease in space
and energy costs resulting from signicantly reducing number of hard links (and
the nerve bers needed to connect them) and replacing them with soft, diusive
pathways.

3.3. The evolutionary advantage of volume transmission

Physical anthropologists considering the possibility of energetic constraints on evo-
lution and exploring where the energy comes from to fuel the large human brain
[54] are puzzled by the fact that the human brain does not use more energy than
the smaller brains of animals of comparable corporal weight. It is likely that the
parts of the human brain that show the greatest size increase over other animals,
such as pre-frontal cortex, may be exactly those parts in which highly nonsynaptic-
based functions have their neuronal representation [26]. For those functions, such as
music appreciation, space-and-energy expensive synaptic neurotransmission may be
largely replaced by VT.
The sensory input and the motor output components of human activities such
as playing a piano concerto are probably highly synaptically organized (although
functions such as vision also have many VT-mediated mechanisms [16]) and do
not dier greatly from comparable functions in non-humans. However, compo-
nents of that activity (playing a piano concerto) are specically human, such
as the musicality and artistic components. These probably involve the speci-
cally human isocortical brain structures, and may not require the high-frequency
(and energetically costly) alternating activation-inactivation of synaptic transmis-
sion. Although direct evidence is lacking, VT is consistent with their modes of
action.
Falk (in [54]) noted: We have to attend to the energetics or were not going to get
selection for a bigger brain. . . . We suggest that VT may play a role in evolution,
providing a mechanism to allow the . . . underlying physical constraints. . . to be
overcome to build an oversize brain. Mitchison [75] suggested that connectivity
appears to be minimized in the brain, and Laughlin et al. [70] noted: . . . neurons,
neural codes and neural circuits will have evolved to reduce metabolic demands.
Functions that are highly VT-mediated may be a basis for the reduced per kilogram
energy requirements of human brains in comparison to the brains of animals of
comparable size.
June 17, 2005 8:59 WSPC/179-JIN 00076

190 Bach-y-Rita

3.4. Recovery from brain damage

My life-long interest in neuroscience led to a mid-career change, from a professorship
in basic science to a resident in physical medicine and rehabilitation. My father had
made a dramatic recovery from a major stroke with a home program developed by
my brother. After my fathers death from a stroke seven years later while mountain
hiking at 9000 feet (he was still working full-time), the autopsy revealed that recovery
had taken place despite very extensive brain damage [1, 8]. The clear demonstration
of restoration of function by means of rehabilitation was an irresistible stimulus for
a career change.
Receptor plasticity, both at synapses and on the cell membrane away from
synapses (reached by VT) may play a major role in the reorganization of function
after brain damage. VT may also be the principal means of neurotransmission in
the noradrenergic system, which is involved in so many activities related to recovery
from brain damage. Both acetylcholine and norepinephrine can provide a state of
excitability consistent with cognition, which is consistent with inhibition of the locus
ceruleus activity during lack of vigilance. These ndings may relate directly to the
results of rehabilitation programs; when vigilance is high and the patient is actively
involved, good results are more likely to be obtained. There may also be a rela-
tionship to functional rehabilitation programs that are based on the interests of the
individual patient, and to the positive results obtained with some home programs.
In these cases, the increased vigilance and participation may lead to greater locus
ceruleus production of noradrenaline. This and other neurotransmitter changes may
also be mechanisms by which psychosocial factors inuence recovery.
Noradrenaline-mediated neurotransmission oers an example of the dierences
between synaptic communication and VT. With VT, the eect of nonjunctional
noradrenaline is likely to be longer lasting than a similar quantity of noradrenaline
released at synapses. Absent such a junction, inactivation of the noradrenaline is
slowed; while the synapse has the full panoply of degradation enzymes and reup-
take mechanisms, nonjunctional receptor sites have few if any of these inactivating
devices [31]. Due to the slow activation and inactivation, there is a reduced need
for nerve action potentials and for neurotransmitter production, transport, and heat
generation. The brain is ecient in its use of energy; thus, the widespread sustained
activation of groups of neurons via diusion of neuroactive substances through the
extracellular uid should be more energy ecient than synaptic activation of those
same neurons [13]. Our modeling studies support this interpretation [4].
Receptor plasticity, both at synapses and on the cell membrane away from
synapses (reached by VT), may play a major role in the reorganization of func-
tion after brain damage. The eects of brain damage are not uniform in the various
neurotransmitter systems. In a rat model, Westerberg et al. [95] studied the eects
of transient cerebral ischemia in the rat hippocampal subelds on excitatory amino
acid receptor ligand binding. They noted that their results demonstrated a lack
of correlation between receptor changes in the early recovery period after ischemia
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 191

and the development of neuronal necrosis in dierent hippocampal regions. They

also noted long-lasting receptor changes in areas considered resistant to an ischemic
insult. Some receptors are down-regulated while others appear to be up-regulated
on the surviving cells. VT may contribute to the survival of partially denervated
neurons and to brain reorganization after brain damage by selective up- and down-
regulation of receptors. These issues have been discussed elsewhere [13, 16, 18, 19, 21],
with regard to brain injury and spinal cord injury.
De Keyser et al. [43] demonstrated an up-regulation of D1 dopamine receptors
in the neocortex of persons whose death was due to a recent unilateral infarct of the
ventral midbrain, producing a unilateral relative dopamine depletion. In autopsies of
patients at 9, 19, and 27 days poststroke, there was a 27% to 37% increase in recep-
tors on the lesioned side. Because the massive brainstem lesion unilaterally destroyed
ascending dopamine bers, it is very likely that the receptor up-regulation was not
related to dopamine synapses. Rather, it is likely that it represents extrasynaptic
VT-related receptors.

3.5. Drug action on the brain

Vizi [91] noted that drugs have diculty reaching the receptors intrasynaptically.
Widely used drugs in clinical practice have been developed or applied on the concept
of presynaptic modulation of neurochemical transmission. He pointed out that the
sensitivity of nonsynaptic receptors is higher, and that they are much more accessible
to drugs. He suggested that diusion neurotransmission may be the primary means
of activation of receptors by externally applied or administered drugs [92].
Drug therapy and rehabilitation may induce functional recovery by inuencing
the aected neurotransmitter systems, and in facilitating the reorganization of brain
functions even when only a small percent of a system (as little as 2%) survives dam-
age [24]. The concept of the eects of drugs on the brain has progressed from a
consideration of the eects of specic drugs on the brain to a consideration of
dierential eects on dierent parts of the brain (our 1956 Science paper [82] may
have been the rst demonstration of the selective regional actions of neuroactive
drugs on the brain), to an understanding of their actions on specic neurotrans-
mitter systems, to the knowledge of their eects on specic intra- and extracellular
mechanisms [15].
VT may be the principal mode of action of psychoactive drugs; other than
conceptual limitations imposed by the present synaptic-dominated model of brain
function, . . . there is no reason to consider that Prozac or any of the drugs used in
psychopharmacology operates exclusively by means of synaptic mechanisms. Accu-
mulating evidence, in fact, suggests the contrary; the primary mechanism may be
types of nonsynaptic diusion neurotransmission [15]. In fact, VT may be the
mechanism by which glia can participate in cognitive functions, since it allows for
neurocommunication in the absence of bers and synaptic connections. Sawa et al.
[83] discuss the role of neuron-glia interactions in the pathogenesis of psychiatric
June 17, 2005 8:59 WSPC/179-JIN 00076

192 Bach-y-Rita

conditions, which may involve a role for VT in negative plasticity leading to a dis-
ease state.

3.6. Extracellular space volume fraction

The extracellular space, which averages more than 20% of the mass of the brain
[78], plays a role in many functions, including VT. In an assembly of brain cells, the
distance between neurons can be reduced by 50% with neuron activity that causes
them to swell [45, 46]. This has an eect on the excitability and metabolism of the
cells by changing the distance between the neurons, which produces, among other
things, changes in ionic concentrations and dynamics [2].
Changes in the size of the extracellular volume fraction may play a role in
membrane excitability in pathologic brain states (e.g., brain damage), and in the
survival of partially denervated neurons during the postinjury period of receptor up-
regulation that can lead to reorganization of brain function by unmasking and other
mechanisms. Under pathologic conditions such as anoxia, the extracellular volume
fraction is reduced [58, 71] and it is also reduced (by up to 50%) in hyperexcitabil-
ity, by changes in the concentration of potassium, and by epileptiform discharges
[45, 46].
Brain cell swelling due to anoxia and brain trauma, leading to a decreased extra-
cellular volume fraction, may aid in the survival of partially denervated neurons
during the postinjury period of receptor up- and down-regulation, which has been
shown to follow brain damage in animal models [95] as well in humans [43]. Those
cells may respond to previously subthreshold stimuli, either synaptically, or by VT,
which generally involves activation of membrane surface receptors. However, it is
also possible that hyperexcitability due to a volume fraction decrease, either inde-
pendently or in combination with excitotoxic activity, may increase secondary cell
death after brain damage [2].

3.7. Computational neuroscience

Hebb imagined cell assemblies as modules comprising enormous numbers of neurons,
each synaptically connected directly to each other. This would lead to factorially
many synaptic connections compared to the number of neurons and virtually all
brain activity would be would be diverted into communications overhead, leaving
practically nothing for actual cognitive activity. This strongly suggests that alter-
native mechanisms such as VT play an important role in the brain.
Hebb [59] noted that integration is fundamentally a question of timing, which has
its eect in the functioning of the cell assembly and the interrelation of assemblies.
He considered them to be diuse, anatomically irregular structures that function
briey as closed systems, and do so only by virtue of the time relations in the ring
of constituent cells. He could not explain the up to half-second delays, which, he
noted, is characteristic of thought. However, delays are a characteristic of the VT
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 193

model of cell assemblies; as noted elsewhere, the initial VT studies revealed delays of
up to 4 seconds, which were considered to be related to multiplexing of polysensory
pontine cells that could have the eect of reducing the numbers of cells needed for
sensory messages [5].
Kercel [66] has pointed out that it is clear that the brain is not a computer,
and crucial brain functions cannot be implemented as a connectionist network.
Semantic ambiguity exists in real-world processes of life and mind. It is inherent
in these processes, and is not a pathological condition that can be resolved by
some sort of process of disambiguation. Semantic ambiguity also exists in high-
order logical abstractions, such as hypersets. Such structures are logically coherent
despite their ambiguity. Thus, it is feasible to rationally investigate a real-world
semantic process, such as the interaction between synaptic communication, glial
communication and VT, by placing the process into a modeling relation, such as
an impredicative logical representation, and gaining novel (albeit qualitative rather
than quantitative) insights about the real-world process by asking questions about
the model.
What is not feasible is serious investigation of such processes by algorithmic com-
putation. Algorithms disallow internal semantics, and specically prohibit ambigu-
ity. Attempting to ignore that prohibition leads to algorithms that deadlock [65].
In other words, in a fundamental manner, the operational constraints of algorithms
dier from the processes of life and mind. Thus, algorithmic descriptions of such
processes are supercial, capturing the incidental syntax but excluding the essential
semantics.
Kercel [64, 66] has further noted that the reductionist paradigm that leads to
connectionism, especially in its modern computational format, claims in princi-
ple to be capable of completely representing any process that occurs in physical
reality. This claim is clearly refuted by its failure to describe inherent semantic
ambiguity, and by its inability to account for the fact that the mind deals eec-
tively with ambiguous sensory cues. The concept of an inherently ambiguous model
that enables us to reason about inherently ambiguous natural processes is gain-
ing interest among logicians and mathematicians [64]. The resulting descriptions
are qualitative and incomplete, but those partial descriptions nevertheless provide
information about relationships and processes that are totally inaccessible to com-
putation.
Kercel [67] has concluded that brain dynamics depends on the interacting dynam-
ics of synaptic, diusive, and glial activities. Observations indicate that synaptic
and diusive activities modify each others morphology, and glial activity modies
both. Synaptic activity modies glial network morphology. Whether diusion mod-
ies glial morphology has not been reported but it is reasonable to expect that
it does. The relationship between these three functions forms a closed-loop hier-
archy of causation in brain function, and the inherent causal ambiguity of such a
hierarchy may account for some of the seemingly bizarre properties of cognitive
function.
June 17, 2005 8:59 WSPC/179-JIN 00076

194 Bach-y-Rita

4. Sensory Substitution and Brain Rehabilitation

4.1. Brain plasticity and the potential for rehabilitation
The need for functional rehabilitation programs that are of interest to the individ-
ual patient has been discussed previously [8, 9, 16, 20]. Such programs are designed
to obtain the active, alert, motivated, and consistent participation of the patient,
thus making maximum use of the patients resources. The patients interest and
dedication add a dimension that may be lost in more formal programs.
Concepts of brain plasticity form the basis for late postacute neurologic rehabil-
itation. Plasticity is a characteristic of our lifelong existence, and thus brain reorga-
nization is possible even many years after brain damage occurs. I have been engaged
in a major eort to counteract the nihilism frequently encountered in the rehabilita-
tion community regarding the potential for late recovery of function. My principal
research models have been late recovery from stroke, late recovery from facial paral-
ysis, and sensory substitution for persons with blindness and other sensory losses.
In the mid-1970s, when newly introduced electronic pong games could be con-
nected to a home television and played by two people controlling individual joy-
sticks, we adapted a pong game for the functional training of upper extremities of
persons with a hemiparetic limb [41]. Three dierent hand pieces were machined to
coincide with the varying grips of the stroke patients for use with a modied exer-
cise track. A patient had to reach forward to 90 of shoulder exion with full elbow
extension to reach balls at the top of the screen. With practice, during the emotion-
ally involving game, the patient ceases to think of arm movements and begins to
think in terms of accomplishing the goal. We noted that the game concept helps to
maintain a high level of interest, enhances motivation, and adds enjoyment to the
hard work of rehabilitation. During a pong game, the patient has an immediate goal
for every movement of the arm. The patient also receives immediate visual feedback
as to the accuracy of the movement. Patients quickly nd themselves absorbed in
playing the game. Some patients prefer playing against the built-in computer or
against their unaected arm. Others enjoy the socialization which develops during
a game with an aide or another patient.
This may have been the rst clinical application of what today would be called
non-immersive virtual reality. We have continued to develop computer assisted moti-
vating rehabilitation (CAMR) programs [34] in various settings, including a recre-
ation rehabilitation program in a beach resort.

4.2. Sensory substitution

Sensory substitution studies were initiated a number of years ago as models of
brain plasticity; persons with congenital sensory loss presented a Jacksonian model
(Hughlings Jackson emphasized the opportunities for discovery oered by the exper-
iments made on the brain by disease [40]). Persons with blindness or other sensory
losses since early infancy do not have one or more of the major aerent inputs, and
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 195

thus have not developed the mechanisms for analyzing information through the lost
system.
A thorough study of persons learning to use a sensory substitution system, with
the information from an articial receptor delivered to the brain through sensory
systems (e.g., tactile) that have remained intact, has oered unique opportunities to
evaluate mechanisms of brain plasticity. It has taken more than 30 years for instru-
mentation technology and neuroscience concepts to mature to the level necessary for
the expected major applications to emerge from those studies, but it now appears
that the stage is set for those to occur.

4.2.1. Vision substation

A person who has suered the total loss of a sensory modality has, indirectly, suered
a brain lesion. In blind persons, about two million bers from the optic nerves
are absent. Absent a modality such as sight, behavior and neural function must
be reorganized. However, blind persons have not necessarily lost the capacity to
see [7], because we do not see with the eyes, but with the brain. In normal sight,
the optical image does not get beyond the retina. From the retina to the central
perceptual structures, the image, now transformed into nerve pulses, is carried over
nerve bers. It is in the central nervous system that pulse-coded information is
interpreted and the subjective visual experience results. It appears to be possible
for the same subjective experience that is produced by a visual image on the retina to
be produced by an optical image captured by an articial eye (a television camera),
when a way is found to deliver the image from the camera to a neural pathway
that can carry it to the brain. Optical images picked up by a television camera are
transduced into a form of energy (vibratory or direct electric stimulation) that can
be mediated by the skin receptors. The visual information reaches the perceptual
levels for analysis and interpretation via somatosensory pathways and structures.
The rest of the process of vision substitution should then depend on experience
and training, and the ability of the subject to have the same control over the image
capture as with eyes; thus, camera movement must be under the control of one of
the subjects motor systems (hand, head movement, or any other). Indeed, we see
[7] have shown that this is possible once the blind person has learned the mechanics.
This includes camera control, zooming, aperture and focus, and the correct inter-
pretation of the eects of camera movement, such as occurs when the camera is
moved from left to right and the image seems to move from right to left. After
sucient training with the tactile vision substitution system (TVSS), our subjects
reported experiencing the image in space, rather than on the skin. They learn to
make perceptual judgments using visual means of analysis, such as perspective, par-
allax, looming and zooming, and depth judgments. Furthermore, many phenomena
associated with vision have to be learned; for example, when viewing a person seated
behind a desk, the partial image of the person must be correctly interpreted as a
complete person with the image of the desk interposed, rather than perceiving just
June 17, 2005 8:59 WSPC/179-JIN 00076

196 Bach-y-Rita

half a person. The subjective experience is comparable, if not qualitatively identical

to, vision, including subjective spatial localization in the three-dimensional world.
Even the visual illusions that have been tested (e.g., waterfall eect) are the same
as vision.
Blind persons were able to use visual means of analysis in two- and three-
dimensional space. Visual illusions and defense reactions to the correctly located
spatial information were obtained. Thus, the brain was shown to be capable of
learning to mediate visual information arriving at the somatosensory cortex, which
apparently does not have the specialized image analysis mechanisms of the visual
cortex.
Many years ago [6], I stated in regard to sensory substitution, The theoreti-
cal basis for the design of the vision substitution system . . . implies a concept of a
brain so malleable that the subjective experience of vision (as well as the quali-
tative and quantitative aerent information necessary for useful vision) could be
obtained through an articial receptor projecting to the cutaneous receptors. Sub-
jective experiences may be the products of a learning process in which aerent inputs
from multiple sources are utilized. . . . In the case of information originating from an
articial receptor, the ability of the brain to extract the data and recreate subjec-
tively the image captured by the articial receptor would depend on the adaptability,
or plasticity, of the brain. There is ample evidence . . . that the brain is capable of
adapting to a variety of extreme conditions. That a successful sensory substitution
system is not presently in use may not be due to limited functional capabilities of the
brain; it may be due to the fact that an articial receptor system has not yet been
constructed to challenge the adaptive capacities of the human brain. Subsequent
tactile vision substitution studies have supported those comments [7, 11, 27].
In 1985 we noted that An understanding of the functional equivalence between
visual and vibrotactile processing would have both basic scientic and practical
implications, the former because it would bear on whether information for the var-
ious perceptual systems ought to be considered modality specic or amodal, and
the latter because the data would suggest the possibilities and constraints for vision
substitution and other prosthetic developments. . . . [28]. Although the early system
was termed a tactile vision substitution system, we have been reluctant to suggest
that blind users of the device are actually seeing. Others [60, 76] have not been so
reluctant, claiming that because blind subjects are being given similar information
to that which causes the sighted to see and are capable of giving similar responses,
one is left with little alternative but to admit that they are seeing (and not merely
seeing).
Kupers et al. [69] used positron emission tomography to study cross-modal plas-
ticity in congenitally blind persons using our TVSS. Participants were trained to
see their tongue in a Snellen orientation detection task. The results showed that the
visual cortex was activated following, but not before, the one week training period.
No visual cortex activation occurred in blindfolded controls. They concluded that
cross-modal plasticity can occur quickly in the adult brain and that the tongue can
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 197

act as a portal to convey somatosensory information to the visual cortex. Years

earlier, we had demonstrated the existence of pathways from the skin to the visual
cortex [77]; these may be unmasked during tactile vision substitution.

4.2.2. Vestibular substation

Persons who have bilateral vestibular damage, such as from the ototoxicity of
gentamicin, experience functional diculties that include postural wobbling (both
sitting and standing), unsteady gait, and oscillopsia. We developed a vestibular
substitution system using a head-mounted accelerometer and an Electro-tactile
Vestibular Substitution System (EVSS). The accelerometer output provides accu-
rate balance information; it is delivered to the brain via an electrotactile array
of stimulators on the tongue. The use of the EVSS produces a strong stabiliza-
tion eect on head-body coordination in subjects with bilateral vestibular dam-
age. Under these conditions, we identied three characteristic and unique head
motion features (drift, sway, long-period perturbations) that consistently appear
in the head-postural behavior of subjects with bilateral vestibular damage, but
are greatly reduced or eliminated with the EVSS [88]. We postulated that in
the absence of the integrated inputs to a normally closed-loop multisensory con-
trol process, an intrinsically unstable system becomes vulnerable to noise (from
both internal and external sources). The recordings show the dramatic eects of
the open-loop control system, which is evidently with the substitute vestibular
input.
Subjects with bilateral vestibular damage using the electrotactile vestibular sub-
stitution system (EVSS) reported feeling normal and stable or having reduced
perceptual noise while using the EVSS and for short periods after removing the
EVSS. Furthermore, the vestibular-cerebellar system evidently has many functions
that are only recently being recognized, such as in dyslexia [35], and so there are
several applications of EVSS that are being explored.
Vestibular substitution subjects demonstrate residual benets, with the ability to
maintain balance even in a totally dark room even after the substitution system has
been disconnected [88]. It is dicult to explain how two twenty-minute vestibular
substitution sessions a day could lead to a therapeutic eect, with recovery persisting
many months after discontinuing use of the substitution system.
But how can this occur? One possibility is suggested by Walls studies; unmask-
ing [93] may play a role in the reorganization. His experiments revealed pathways
that exist in the normal state, but do not appear to function until unmasked by
injury or temporary conduction block. Comparably, the existence of weak non-visual
inputs to cells of the cat primary visual cortex had previously been demonstrated;
the responses to light were tightly grouped around a mean of 33 milliseconds in
latency while those to sound and pinprick were spread out around means of 63
(sound) and 70 (pinprick) milliseconds [7]. Further, these responses to sound and
pinprick were less synaptically secure, being less resistant to barbiturate narcosis
June 17, 2005 8:59 WSPC/179-JIN 00076

198 Bach-y-Rita

than the visual responses. Studies in adult congenitally blind persons reveal that
the non-visual inputs to the visual cortex become prominent [94]. These more ten-
uous pathways may be the type of masked pathways that are unmasked following
neural lesion, provided that there is an appropriate rehabilitation or substitution
program, and that there is the functional demand and the motivation to obtain the
increased function.

5. Beyond Sensory Substitution Human-in-the-Loop

As sensory substitution via the tonguebrain-machine interface has worked for var-
ious sensory modalities, it is likely that almost any electronically generated data
stream can be coupled into the human nervous system through the brain-machine
interface, and can then be experienced by the user as a Gestalt, or direct perception.
Applications could include the monitoring of data streams generated by industrial,
commercial, and military processes [29]. In a eld such as industrial control, incor-
porating humans-in-the-loop should greatly increase speed and eciency, thereby
reducing fatigue.
Instead of reading meters and readouts, interpreting results and pondering what
they mean about the process, the operator would experience the process by direct
perception. For example, the ability to determine directly and inexpensively whether
or not the product meets specications is an enormous problem for the steel industry
(and for virtually all other manufacturers). A thickness gauge in a tube-piercing mill
produces a thickness prole consisting of approximately 3000 numbers per tube.
Interpreting the numbers to tell whether the tube is within specication is dicult
and tedious. Incorporating humans-in-the-loop, the operator would experience the
tube passing through the gauge in much the same way as one experiences the feeling
of sliding the nger over a surface. The experience might feel smooth or it might
feel rough or somewhere in between. The well-trained operator could distinguish
directly from experiencing the perception whether the tube is smooth enough to
meet standards.
Direct perception could lead to automating workers into the process, instead of
the more conventional strategy of automating them out. It should also complement
the present eort to engineer for fully autonomic systems and lead to greater use of
systems with the human engineered in-the-loop [30, 68].

6. Why the Delay in Acceptance?

The tactile sensory substitution project began in 1963. Initially it was impossible
to fund. With junk-pile equipment (e.g., a discarded dental chair and a discarded
bulky tripod mounted camera) excellent results were obtained, and published in
a Nature paper in 1969 [27]. We proved tactile vision substitution beyond doubt.
Nevertheless, it has taken a further 35 years to rmly establish it in the neuroscience
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 199

community. This has been the source of frustration, but at least now it does not
shock people or worse, make them laugh in disdain. More signicantly, we are now
receiving substantial and consistent research funding.
But why has it taken so long? Clearly it was far out of the mainstream of science.
In addition, it posed a very simple solution to a major problem. There are very
extensive and complex mechanisms involved in vision, but we were oering a simple
low-resolution visual image through the skin, and it works. It is an example of the
tremendous capacity of the brain to extract information from even the simplest
input. An example is the enormously useful information gathered by a blind person
entering a room using a long cane. A single point swept sequentially provides the
information to create a spatial knowledge of the room and contents.
I have previously noted that blind subjects who have trained with the TVSS
demonstrate perceptual equivalence between and across modalities [16], and that
this is also frequently noted under other circumstances. Gibson [55] noted that re
is the same whether the information has been obtained by hearing, feeling, looking,
or smelling. There is a common aspect of perceptual activity that permits one to
utilize information from several channels in such a way that invariant properties of
objects are extracted.
As learning progresses, the information extraction processes become more and
more automatic and unconscious. Miller [74] considered that a chunking phe-
nomenon allows the number of bits per chunk to increase. A blind subject looking
at a display of objects must initially consciously perceive each of the relative factors
such as the perspective of the table, the precise contour of each object, the size and
orientation of each object and the relative position of parts of each object to others
nearby. With experience, information regarding several of these factors is simultane-
ously gathered and evaluated. The increased information transfer through a sensory
substitution system can be interpreted in terms of Millers chunking model. The
highly complex visual input can thus be reduced by selective processes to man-
ageable proportions, allowing the input to be mediated by the somesthetic system
or, in Gibsonian [55] terms, the subject learns to extract the relevant information.
Many relevant studies have been summarized in Dennetts 1990 book [44]. He
noted that There are perceptual analysis tasks, such as speech perception, that
would be beyond the physical limits of the brains machinery if it didnt utilize
ingenious anticipatory strategies that feed on redundancies in the input. The world
provides an inexhaustible deluge of information bombarding our senses and when we
concentrate on how much is coming in, or continuously available, we often succumb
to the illusion that all of it must be used at all time. However, our capacities to use
information and our epistemic appetites are limited. If our brains can just satisfy
all our particular epistemic hungers as they arise . . . we will never be able to tell . . .
that our brains are provisioning us with less than everything that is available in the
world. A conscious mind is an observer, who takes in a limited subset of all the
information . . . available at a particular . . . continuous sequence of times and places.
June 17, 2005 8:59 WSPC/179-JIN 00076

200 Bach-y-Rita

. . . all varieties of perception indeed, all varieties of thought or mental activity

are accomplished in the brain by parallel, multitrack processes of interpretation and
elaboration of sensory inputs. . . Feature detections or discriminations only have to
be made once.
Such comments aid in the understanding of the capacity of a low-resolution
sensory substitution system to provide the information necessary for complex image
interpretation. Furthermore, even the stated inadequacies of the skin (e.g., poor two-
point resolution) do not appear as serious barriers to eventual high performance.

7. Closing Comments
In April 2004, I celebrated my 70th birthday shortly after I was diagnosed with a
stage-4 lung cancer. Both of these events are inducements to reviewing my lifes
work. What I would like to see happening in the near future is the birth of a for-
mal eld of Theoretical Neuroscience which would allow the evolution of concepts
of brain function and lead to applications in the health elds. Thus, as experimen-
tal evidence becomes available and challenging the concepts being debated within
Theoretical Neuroscience, old ideas would be displaced much more rapidly than
the 50 or more years that it now takes. Mechanisms, such as VT, which may very
well compose the majority of neuro-communication in the brain and play important
roles in brain energetics, evolution of species, brain plasticity and recovery of func-
tion after brain damage or sensory loss, would be recognized and integrated into
new insights and practical applications in learning and recovery. Mechanisms such
as those related to motivation and consciousness that have not been evaluated by
available technology, would be acknowledged as legitimate areas for serious study
by neuroscientists. The dialogue between humanistic and scientic approaches to
human communication would increase from its present levels. A eld of Theoretical
Neurosciences would bridge the gap between the laboratory and the public and serve
to make the capacities of healing and reorganization of the brain, universally known
and immediately useable.

Acknowledgments
These studies were supported by an SBIR2 and an RO1 from the NEI, and an SBIR2
from the NIDCD, National Institutes of Health.

References
[1] Aguilar MJ, Recovery of motor function after unilateral infarction of the basis pontis,
Report of a case, Amer J Physical Med 48:279288, 1969.
[2] Aiello GL, Bach-y-Rita P, Brain cell microenvironment eects on neuron excitability
and basal metabolism, Neuroreport 8:11651168, 1997.
[3] Aiello GL, Bach-y-Rita P, The cost of an action potential, J Neurosci Meth 103:145
149, 2000.
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 201

[4] Aiello GL, Bach-y-Rita P, Hebbian brain cell-assemblies: Nonsynaptic neurotrans-

mission, space and energy considerations, in Dagli CH, Buczak AL, Embrechts MJ,
Ersoy O, Kercel SW (eds.), Intelligent Engineering Systems through Articial Neural
Networks Volume 11, ASME Press, New York, pp. 441447, 2001.
[5] Bach-y-Rita P, Convergent and long-latency unit responses in the reticular formation
of the cat, Exp Neurol 9:327344, 1964.
[6] Bach-y-Rita P, Sensory plasticity: Applications to a vision substitution system, Acta
Neurol Scand 43:417426, 1967.
[7] Bach-y-Rita P, Brain Mechanisms in Sensory Substitution, Academic Press, New York,
1972.
[8] Bach-y-Rita P, Brain plasticity as a basis for therapeutic procedures, in Bach-y-Rita P
(ed.), Recovery of Function: Theoretical Considerations for Brain Injury Rehabilitation,
Hans Huber, Berne Switzerland, pp. 225263, 1980.
[9] Bach-y-Rita P, Brain plasticity as a basis of the development of rehabilitation proce-
dures for hemiplegia, Scand J Rehabil Med 13:7383, 1981.
[10] Bach-y-Rita P, Brain plasticity, in Goodgold J (ed.), Rehabilitation Medicine, CV
Mosby Co., St. Louis MO, pp. 113118, 1988.
[11] Bach-y-Rita P, Physiological considerations in sensory enhancement and substitution,
Eur Med Phys 25:107128, 1989.
[12] Bach-y-Rita P, Thoughts on the role of volume transmission in normal and abnormal
mass sustained functions, in Fuxe K, Agnati LF (eds.), Volume Transmission in the
Brain, Raven Press, New York, pp. 489496, 1991.
[13] Bach-y-Rita P, Neurotransmission in the brain by diusion through the extracellular
uid: A review, Neuroreport 4:343350, 1993.
[14] Bach-y-Rita P, The brain beyond the synapse: A review, Neuroreport 5:15531557,
1994.
[15] Bach-y-Rita P, Psychopharmacologic drugs: mechanisms of action (Letter), Science
264:642644, 1994.
[16] Bach-y-Rita P, Nonsynaptic Diusion Neurotransmission and Late Brain Reorganiza-
tion, Demos Publications, New York, 1995.
[17] Bach-y-Rita P, Conservation of space and energy in the brain, Restor Neurol Neuros
10:13, 1996.
[18] Bach-y-Rita P, Nonsynaptic diusion neurotransmission and some other emerging con-
cepts, P W Pharmacol Soc 41:211218, 1998.
[19] Bach-y-Rita P, Theoretical aspects of sensory substitution and of neurotransmission-
related reorganization in spinal cord injury, Spinal Cord 37:465474, 1999.
[20] Bach-y-Rita P, Conceptual issues relevant to present and future neurologic reha-
bilitation, in Levin H, Grafman J (eds.), Neuroplasticity and Reorganization of
Function after Brain Injury, Oxford University Press, New York, pp. 357379,
2000.
[21] Bach-y-Rita P, Theoretical and practical considerations in the restoration of functions
after stroke, Top Stroke Rehabil 8:115, 2001.
[22] Bach-y-Rita P, Volume transmission and brain plasticity, Evol Cognition 8:115122,
2002.
[23] Bach-y-Rita P, Brain damage, recovery from, in Ramachandran VS (ed.), Encyclopedia
of the Human Brain Vol 1, Academic Press, New York, pp. 481491, 2002.
June 17, 2005 8:59 WSPC/179-JIN 00076

202 Bach-y-Rita

[24] Bach-y-Rita P, Is it possible to restore function with two percent surviving neural
tissue? J Integr Neurosci 3:36, 2004.
[25] Bach-y-Rita P, Aiello GL, Nerve length and volume in synaptic versus diusion neuro-
transmission: A model, Neuroreport 7:15021504, 1996.
[26] Bach-y-Rita P, Aiello GL, Brain energetics and evolution, Brain Behav Sci 24:280,
2001.
[27] Bach-y-Rita P, Collins CC, Saunders F, White B, Scadden L, Vision substitution by
tactile image projection, Nature 221:963964, 1969.
[28] Bach-y-Rita P, Hughes B, Tactile vision substitution: Some instrumentation and per-
ceptual considerations, in Warren D, Strelow E (eds.), Electronic Spatial Sensing for
the Blind, Martinus-Nijho, Dordrecht the Netherlands, pp. 171186, 1985.
[29] Bach-y-Rita P, Kercel SW, Sensori-motor coupling by observed and imagined move-
ment, Intellectica 35:287297, 2002.
[30] Bach-y-Rita P, Kercel SW, A novel strategy for coupling data into the human nervous
system, in OHara JM, Johnson G (eds.), Proceedings of the Fourth International Topi-
cal Meeting on Nuclear Plant Instrumentation, Control and Human-Machine Interface
Technology, American Nuclear Society, LaGrange Park IL, 2004.
[31] Bach-y-Rita P, Smith CUM, Comparative eciency of volume and synaptic transmis-
sion in the coerulean system: Relevance to neurologic rehabilitation, Scand J Rehabil
Med 25:36, 1993.
[32] Bach-y-Rita P, Wicab-Bach-y-Rita E, Biological and psychosocial factors in recovery
from brain damage in humans, Can J Psychology 44:148165, 1990.
[33] Bach-y-Rita P, Wicab-Bach-y-Rita E, Hope and active patient participation in the
rehabilition environment, Arch Phys Med Rehabil 71:10841085, 1990.
[34] Bach-y-Rita P, Wood S, Leder R, Parades O, Bahr D, Wicab-Bach-y-Rita E, Murillo N,
Computer-assisted motivating rehabilitation (CAMR) for institutional, home, and edu-
cational late stroke programs, Top Stroke Rehabil 8:110, 2002.
[35] Beaton AA, Dyslexia and the cerebellar decit hypothesis, Cortex 38:479490,
2002.
[36] Beaudet A, Descarries L, The monoamine innervation of rat cerebral cortex: Synaptic
and nonsynaptic axon terminals, Neuroscience 3:851860, 1978.
[37] Broca P, Localisations des fonctions cerebrales. Siege de la faculte du langage articule,
Bull Soc Anthropol 4:200208, 1863.
[38] Cajal SRy, Recollections of My Life, MIT Press, Cambridge MA, 1989.
[39] Chow KL, Stewart DL, Reversal of structural and functional eects of long-term visual
deprivation in cats, Exp Neurol 34:409433, 1972.
[40] Clarke E, OMalley CD, The Human Brain and Spinal Cord, University of California
Press, Berkeley, 1968.
[41] Cogan A, Madey J, Kaufman W, Holmlund G, Bach-y-Rita P, Pong game as a rehabil-
itation device, in Warren, CG (ed.), Proceedings of the Fourth Annual Conference on
Systems and Devices for the Disabled, University of Washington School of Medicine,
Seattle WA, pp. 187188, 1977.
[42] Colotla VA, Bach-y-Rita P, Shepherd Ivory Franz: His contributions to neuropsychology
and rehabilitation, Cogn Aect Behav Neurosci 2:141148, 2002.
[43] De Keyser JD, Ebinger G, Vauquelin G, Evidence for a widespread dopaminergic inner-
vation of the human cerebral neocortex, Neurosci Lett 104:281285, 1989.
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 203

[44] Dennett DC, Consciousness Explained, Penguin Books, London, 1993.

[45] Dietzel I, Heinemann U, Hofmeier G, Lux HD, Transient changes in the size of the
extracellular space in the sensorimotor cortex of cats in relation to stimulus-induced
changes in potassium concentration, Exp Brain Res 40:432439, 1980.
[46] Dietzel I, Heinemann U, Dynamic variations of the brain cell microenvironment in
relation to neuronal hyperactivity, Ann NY Acad Sci 481:7286, 1986.
[47] Edelman, GM, Bright Air, Brilliant Fire, Basic Books, New York, 1992.
[48] Franz SI, Scheetz ME, Wilson AA, The possibility of recovery of motor function in
long-standing hemiplegia, J Amer Med Assoc 24:21502154, 1915.
[49] Freeman WJ, Premises in neurophysiological studies of learning, in Lynch G,
McGaugh JL, Weinberger NM (eds.), Neurobiology of Learning and Memory, Guilford
Press, New York, pp. 231234, 1984.
[50] Freeman WJ, Predictions on neocortical dynamics derived from studies in paleocortex,
in Basar E, Bulloch TH (eds.), Induced Rhythms in the Brain, Birkhauser, Boston,
pp. 183199, 1992.
[51] Freeman WJ, Tutorial on neurobiology: From single neurons to brain chaos, Int J
Bifurcat Chaos 2:451482, 1992.
[52] Freeman WJ, Societies of Brains: A Study in the Neuroscience of Love and Hate,
Lawrence Erlbaum Associates, New York, 1995.
[53] Fuxe K, Agnati LF, Volume Transmission in the Brain: Novel Mechanisms for Neural
Transmission, Raven Press, New York, 1991.
[54] Gibbons A, Solving the brains energy crisis, Science 280:13451347, 1998.
[55] Gibson JJ, The Senses Considered as Perceptual Systems, Houghton Miin, Boston,
1966.
[56] Gilbert CD, Hirsch JA, Weisel TN, Lateral interactions in the visual cortex, Cold Spring
Harb Sym 55:663677, 1990.
[57] Gilbert CD, Wiesel TN, Receptive eld dynamics in adult primary visual cortex, Nature
356:150152, 1992.
[58] Harreveld AV, Khattab FI, Changes in cortical extracellular space during spreading
depression investigated with the electron microscope, J Neurophysiol 30:911929, 1967.
[59] Hebb DO, The Organization of Behavior, John Wiley & Sons, New York, 1949.
[60] Heil J, Perception and Cognition, University of California Press, Berkeley, 1983.
[61] Herkenham M, Mismatches between neurotransmitter and receptor localizations in
brain: Observations and implications, Neuroscience 23:138, 1987.
[62] Horrobin D, Discouraging hypotheses slows progress, The Scientist 4:1314, 1990.
[63] Hubel DH, Wiesel TN, The period of susceptibility to the physiological eects of uni-
lateral eye closure in kittens, J Physiol London 206:419436, 1970.
[64] Kercel SW, Endogenous causes bizarre eects, Evol Cognition 8:130144, 2002.
[65] Kercel SW, Softer than soft computing, in Kercel SW (ed.), Proceedings of the 2003
IEEE International Workshop on Soft Computing in Industrial Applications, IEEE,
Piscataway NJ, pp. 2732, 2003.
[66] Kercel SW, The endogenous brain, J Integr Neurosci 3:6184, 2004.
[67] Kercel SW, The role of volume transmission in an endogenous brain, J Integr Neurosci
3:718, 2004.
[68] Kercel SW, Manges WW, Reber AS, Wiring a human-in-the-loop, in OHara JM, John-
son G (eds.), Proceedings of the Fourth International Topical Meeting on Nuclear Plant
June 17, 2005 8:59 WSPC/179-JIN 00076

204 Bach-y-Rita

Instrumentation, Control and Human-Machine Interface Technology, American Nuclear

Society, LaGrange Park IL, 2004.
[69] Kupers R, Sampaio E, Moesgaard S, Gjedde A, Ptito M, Activation of visual cortex
by electrotactile stimulation of the tongue in early-blind subjects, Neuroimage 19:S65,
2003.
[70] Laughlin SB, de Ruyter van Stevenick RR, Anderson JC, The metabolic cost of neural
information, Nat Neurosci 1:3641, 1998.
[71] Lipton SA, Kater SB, Neurotransmitter regulation of neuronal outgrowth, plasticity
and survival, Trends Neurosci 12:265270, 1989.
[72] Luria AR, Restoration of Function After Brain Damage, Macmillan, New York, 1963.
[73] Merton RK, The self-fullling prophecy, in Merton RK (ed.), Social Theory and Social
Structure, Free Press, New York, pp. 475490, 1968.
[74] Miller GA, The magical number seven, plus or minus two: Some limits on our capacity
to process information, Psychol Rev 63:8197, 1956.
[75] Mitchison G, Axonal trees and cortical architecture, Trends Neurosci 15:122126, 1992.
[76] Morgan MJ, Molyneuxs Question, Cambridge University Press, Cambridge, 1977.
[77] Murata K, Cramer H, Bach-y-Rita P, Neuronal convergence of noxious, acoustic and
visual stimuli in the visual cortex of the cat, J Neurophysiol 28:12231239, 1965.
[78] Nicholson C, Volume Transmission in the year 2000, in Volume Transmission Revis-
ited Agnati L, Fuxe K, Nicholson C, Sykova E, [Progress in Brain Research Vol. 125]
Elsevier, Oxford, pp. 437446, 2000.
[79] Nicholson C, Phillips JM, Ion diusion modied by tortuosity and volume fraction in
the extracellular microenvironment of the rat cerebellum, J Physiol London 321:225
257, 1981.
[80] Nieuwenhuys R, Chemoarchitecture of the Brain, Springer, Berlin, 1986.
[81] Nieuwenhuys R, Comparative aspects of volume transmission, with sidelight on other
forms of intercellular communication. In Volume Transmission Revisited Agnati L, Fuxe
K, Nicholson C and Sykova E, [Progress in Brain Research Vol. 125] Elsevier, Oxford,
pp. 49128, 2000.
[82] Olds J, Killam KF, Bach-y-Rita P, Self-stimulation of the brain used as a screening
method for tranquilizing drugs, Science 124:265266, 1956.
[83] Sawa A, Pletnikov MV, Kamiya A, Neuron-glia interactions clarify genetic-
environmental links in mental illness, Trends Neurosci 27:294297, 2004.
[84] Shepherd GM, Foundations of the Neuron Doctrine (History of Neuroscience, No. 6),
Oxford University Press, Oxford, 1991.
[85] Sherrington CS, The Integrative Action of the Nervous System, Yale University Press,
New Haven CT, 1947.
[86] Sholl DA, The Organization of the Cerebral Cortex, John Wiley & Sons, New York,
1956.
[87] Taub E, Motor behavior following deaerentiation in the developing and motorically
mature monkey, in Herman R, Grillner S, Ralston HJ, Stein PSD, Stuart D (eds.),
Neural Control of Locomotion, Plenum, New York, pp. 675705, 1976.
[88] Tyler ME, Danilov Y, Bach-y-Rita P, Closing an open-loop control system: Vestibular
substitution through the tongue, J Integr Neurosci 2:159164, 2003.
[89] Vizi ES, Presynaptic modulation of neurochemical transmission, Prog Neurobiol
12:181290, 1979.
June 17, 2005 8:59 WSPC/179-JIN 00076

Emerging Concepts of Brain Function 205

[90] Vizi ES, Non-synaptic modulation of transmitter release: Pharmacological implications,

Trends Pharmacol Sci 2:172175, 1980.
[91] Vizi ES, Nonsynaptic inhibitory signal transmission between axon terminals: Physiolog-
ical and pharmacological evidence, in Fuxe K, Agnati LF (eds.), Volume Transmission
in the Brain, Raven Press, New York, pp. 8996, 1991.
[92] Vizi ES, Role of high anity receptors and membrane transporters in nonsynap-
tic communication and drug action in the central nervous system, Pharmacol Rev
52:6389, 2000.
[93] Wall PD, Mechanisms of plasticity of connection following damage in adult mammalian
nervous systems, in Bach-y-Rita P, (ed.), Recovery of Function: Theoretical Consider-
ations for Brain Injury Rehabilitation, Hans Huber, Berne Switzerland, pp. 91105,
1980.
[94] Wanet-Defalque MC, Veraart C, DeVolder A, Metz R, Michel C, Dooms G, Gonet
A, High metabolic activity in the visual cortex of early blind human subjects, Brain
Res 446:369373, 1988.
[95] Westerberg E, Monaghan DT, Kalimo H, Cotman CW, Wieloch TW, Dynamic changes
of excitatory amino acid receptors in the rat hippocampus following transient cerebral
ischemia, J Neurosci 9:798805, 1989.
[96] Zoli M, Jansson A, Sykova E, Agnati LF, Fuxe K, Volume transmission in the CNS and
its relevance for neuropsychopharmacology, Trends Pharmacol Sci 20:142150, 1999.