When discussing this question it can be quite misleading to assume that the brain is entirely under our volitional control. In fact many parts of the human brain, such as the cerebellum, the mid brain and the pons, control essential bodily functions such as breathing and other semi automated functions which we are mostly unaware of. What most people are possibly hinting at, in suggesting that we only use ten per cent of our brain, is the role of the cerebrum and its conscious control over our actions. Then again, some psychologists would maintain that the feeling of being in control is an illusion, a way of rationalizing our actions after the event.
Depending on whether we are sleeping, or we are awake and in various levels of activity, we use an ever increasing percentage of our brain which is generally in excess of ten per cent. By way of an analogy we can compare the brain to a generator which is always working to maintain power to our electrical circuits. Degrees of mental activity can be compared to the numbers of lights which are switched on to tap into the generated electrical current. Switching off all the lights still leaves the generator humming away maintaining the power potential. The use of an electroencephalograph, EEG, can actually measure this background hum in the brain.
Perhaps the question ought to be rephrased to relate to what percentage of conscious control we have over activities primarily controlled by the cerebrum. It could be argued that when we are asleep we must be using hardly any of the cerebrum's capacity. However magnetic resonance imaging, MRI, research would tend to contradict this argument, demonstrating that there is a veritable hive of activity taking place even during so-called unconscious states. Some researchers would maintain that we are never truly not conscious as long as we are alive in the sense that we are not brain dead.
In order to understand more fully the role of the cerebrum in conscious activities such as thinking and decision making we need to examine these processes in greater detail. Techniques such as MRI and PET, positron emission topography, reveal extensive downward projections from cortex to thalamus. A consensus view has emerged in which reverberatory feedback between the thalamus and pyramidal cells in the cortex are thought to provide the neural correlate of consciousness. Electrophysiological recordings, EEG, demonstrate coherent firing of thalamo-cortical loops with frequencies from slow EEG frequencies 2-12Hz (cycles per second) to rapid gamma oscillations in the 40 Hz range and upwards. Coherent gamma frequency thalamo-cortical oscillations (40 Hz) are suggested to mediate temporal binding of conscious experience. Some discussion surrounds whether coherence originates in the thalamus or resonates in the cortical networks but reverberatory loops remain a prevalent classical viewpoint as to the source of human consciousness.
It is thought that thalamo-cortical loops are selected and maintained by axonal-dendritic chemical synaptic transmission; passing signals across the gaps between neurons synapses. The terminal axon of neuron A releases a neurotransmitter (chemical substance) which then binds to a post-synaptic receptor on a dendritic spine on neuron B. This changes the local dendritic membrane potential (electrical charge) on neuron B which interacts with neighboring spines and dendrites until neuron B reaches the threshold for firing; passing on the signal. Axonal depolarizations firings or spikes travel down the axon of neuron B and result in the release of neurotransmitter vesicles from the pre-synaptic axon terminal into another synapse. The neurotransmitter binds to post-synaptic dendritic spine receptors on neuron C and this process continues onwards to other neurons. The idea is that each neuron provides multiple inputs and analogue (continuous) processing in dendrites until a threshold is reached for axonal firings that produce a single digital (on/off) output, a bit or unit of information in the brain.
Changes to the chemical synaptic network patterns between neurons are presumed to correlate with particular mental states, changing dynamically according to the relative strengths of the chemical synapses that join them. This synaptic strength shapes the network dynamics and is responsible for cognitive effects such as learning by selecting particular neural network patterns over others, possibly because they correlate with existing networks where they fit-in and add to the individual's store of knowledge. Dendritic-dendritic processing is another possibility, including processing among dendrites on the same neuron. Eccles points to dendritic arborizations (tree-like branching) of pyramidal cells in the cerebral cortex as the locus of conscious processing. Crick suggests that the mechanical dynamics, structurally mediated interactions, of the dendritic spines are essential to higher-level processes including consciousness.
In addition to chemical synaptic junctions in the brain there is also a prevalence of gap junction connections. Neurons and glia (cells between neurons) are interconnected by electrotonic gap junctions, which are window-like portholes between adjacent neural processes, axon-dendrite, dendrite-dendrite, dendrite-glial cell. Cytoplasm flows through this 4 nm (nanometer) gap between the two processes and the cells are synchronously coupled behaving in some instances like a single neuron. Gap junctions may be considered to be more primitive than chemical synapses and essential only for embryological development but they remain active throughout adult life and account for about fifteen percent of all brain connections.
Gap junctions could also play a significant role in the spread of a quantum synchronous state, rather than a classical synchronous state, between neurons and glia. Electron tunneling which is known to enlarge quantum states and has been demonstrated to occur over a distance of five nanometers, could easily bridge the four-nanometer gap junctions between adjacent neurons. Critically such a spread would occur almost instantaneously, a peculiarity of quantum physics which distinguishes it from classical physics. Specific intracellular organelles, in some instances a mitochondria (cell's energy sources), have been discovered in dendrites immediately adjacent to gap junctions. These dendritic lamellar bodies are tethered to small cytoskeleton proteins that are fastened to microtubules. The mitochondria could provide free electrons for the process of tunneling, acting as a sort of Josephson junction (quantum link) between cells for the spread of a cytoplasmic quantum state.
In the accepted classical model axonal firings resulting in neurotransmitter release are the basic currency of information. However only about fifteen percent of axonal firings result in neurotransmitter release and the probability of such release occurring appears to be random. Either this is a huge wastage of effort exacerbated by unpredictable responses or other factors must be involved. Local factors like the cytoskeleton could be modulating the whole process bringing into play possible quantum level influences.
Electrophysiological recordings (EEG) have provided much of what is known about neuronal behavior within the brain. These recording typically include a great deal of background drift or noise which electro physiologists attempt to eliminate by averaging across several recordings. The actual source of this noise has not been confirmed and there are those who question whether it is noise at all since it appears to be correlated across brain regions. It has been conjectured by some researchers, Zohar and Marshall included, that this phenomenon represents some as yet unrecognized form of signaling, or possibly it is simply the hum of the generator; the brain constantly at work.
Conventional approaches to cognition and consciousness propose synapses and membrane proteins to be the bottom rungs of the hierarchical processes involved in the brain. Membrane proteins exist in a generally unstable environment and are quite short-lived having to be reformed within a matter of hours or days. According to Hameroff, they are maintained by a system of axoplasmic transport (along the axon of the neuron) in which membrane proteins are synthesized in the cell body of the neuron and supplied to the synapse by microtubules. It is likely that the microtubules act as more than just convenient connections for the transport of materials since it is known that they also establish new synapses and are responsible for regulating existing synaptic efficacy. It is possible that the sensitivity of synapses depends on the cytoskeleton microtubules whose functioning is also related to quantum level processes (switching between different conformational quantum states of tubulin).
The conventional approach to intracellular information processing assumes that this primarily occurs by diffusion of soluble biomolecules, ionic flux etc. This process is by necessity slow and haphazard particularly across neural processes. Alternatively the cytoskeleton microtubules offer a solid-state' quantum system well suited for rapid information processing. In unicellular organisms such as Paramecium, whose survival depends on reacting quickly to changes in the environment, threats and opportunities, the slow diffusion model would appear to be altogether untenable. However unicellular organisms possess microtubules made of tubulin in the cytoskeleton of their cells which would support the argument for quantum level processes. Ergo, if quantum processes are involved in information processing in single celled organisms like Paramecium it seems reasonable to propose that they could also be involved in similar ways in the cells of multicellular organisms like humans.
With so much activity within the brain happening all the time, it would appear quite illogical to suggest that humans use only ten per cent of its total capacity. It would be more accurate to suggest that sometimes we use only a small percentage of those parts of the brain which appear to be under our conscious control. The brain is similar to the body's muscular system in that it requires exercise to keep it working efficiently. Failure to provide adequate stimulation for the brain literally causes it to lose the number of synaptic connections which connect brain cells together. Once synaptic connections are lost there is a tendency for that part of the brain to atrophy producing dementia and other conditions often associated with aging. This reminds one of the common adage, "use it or lose it".