We propose that the nucleus of a single neuron regulates the synapses connected to a unique group of follower cells via collaterals for the purpose of creating one unit of consciousness (Tehovnik, Hasanbegović, Chen 2024). The following empirical points support the proposal:

(1) We know that in the neocortex of humans there are approximately 1.6 x 10^10 neurons with a total synaptic count of about 1.6 x 10^14 synapses such that each neuron on average has 1 x 10^4 synapses (Tehovnik, Hasanbegović, Chen 2024), some of which may be to the same follower cell and some to a different follower cell. Assume for a moment that a single neocortical neuron innervates 5,000 neocortical follower neurons (perhaps an overestimation, but this is not the point of the argument). This represents one unit of consciousness for we know that a single neuron can be conditioned volitionally using a brain machine interface paradigm (Prsa, Huber et al. 2017) and that such conditioning requires severe attentional effort, as evidenced by human subjects learning to control brain-machine-interface devices (Bublitz et al. 2018). Also, it is known that consciousness of the elements supporting language or mathematics can be suppressed/erased by focusing electrical currents at specific points within the neocortical grey-matter sheet (Penfield and Roberts 1966; Ojemann 1991). Here is the first example of a quasi-unit of consciousness being disrupted, albeit more than one neuron was likely activated in these studies. Nevertheless, we now have a methodology to activate a single neuron extracellularly in primates (see Fig. 7 of Tehovnik et al. 2009) and intracellular studies have activated single neurons in animals that can be detected by the animal (Houweling and Brecht 2008).

(2) The approximate duration of a thought can be a fraction of a second up to several seconds before transitioning into another thought, based on metrics from binocular rivalry and visual experiments (Dwarakanathetal, Logothetis 2023; Schiller and Tehovnik 2015). Also, it seems that the trigger for a transition in conscious thought is a burst of theta activity (Dwarakanathetal, Logothetis 2023), which has a subcortical origin perhaps at the level of the septum (Buzsáki 2006). Since we are on the topic of triggers for transitions in consciousness, we need to dispense with the idea that midbrain dopamine is the trigger, given that all movement and thinking can be arrested in severe Parkinsonism to the point of evoking death (Fuentes per. com 2011; Sacks 1976, 2012). The firing properties of dopaminergic fibres destined for neocortex is such that there is a restrictive range of firing since the fibres are unmyelinated such that frequency of firing remains at about 1 Hz (Tehovnik, Hasanbegović, Chen 2024). This restricted range reduces the dopamine system to at best being an ON and OFF switch. This propensity makes L-DOPA and other recovery drugs so effective in restoring total functionality to Parkinson’s patients even though the recovery can be short-lived due to adaptation (Sacks 1976, 2012). Dopamine is more of a lubricant that facilitates the smooth transition of consciousness such that the rate of conscious thought can be affected by the amount of cocaine or amphetamine one ingests, which potentiates the dopamine system (Tehovnik, Hasanbegović, Chen 2024). Too much of either can produce schizophrenic symptomatology, which means such agents need to be titrated for optimal effectiveness. Also, while one is depressed it is more likely than not that the dopamine system is under-performing and pushing a subject toward a Parkinsonian state, thereby slowing down transitions in consciousness. This explains why depression often evokes self-chosen immobility. To conclude, theta is the main trigger for the transitions in consciousness, with dopamine playing a minor role despite all the speculation of its importance in cognition and motor control (e.g., Matsumoto, Graybiel et al. 1999).

(3) The idea that consciousness is a stream (James 1890) and that it is continuous during waking state (Chomsky 2012) is not in dispute for one just needs to introspect on this point to be convinced. How does one transition into a second conscious state to start the streaming process after opening one’s eyes in the morning? Two points: first we believe that the conscious process (outside of learning something new) is triggered by signals emanating from the cerebellum (Ikeda and Kimura 1994; Libet 1984), which stores all the information from past consolidations covering an entire lifespan to the present day (with a storage capacity of 2.8 x 10^14 bits, Tehovnik, Hasanbegović, Chen 2024). Once triggered by accessing an initial consciousness unit, this unit is connected to other units based on previous learning (Hebb 1949, 1960, 1968); this connectedness is what makes consciousness coherent and familiar to us. This process is universal at least across all vertebrates, which contain both a telencephalon/hippocampus (a neocortical homologue) and a cerebellum (Cisek 2019; Herculano-Houzel 2009; Murray, Wise et al. 2017). The concatenation process of consciousness at the level of neocortex is also very familiar to us, particularly when consciousness is actualized in the form of a motor, a verbal, or a musical sequence. Actualization requires participation from the cerebellum and the motor nuclei that reside in the brain stem and spinal cord. In the absence of engaging the motor system (i.e., the cerebellum and the motor nuclei) a subject is reduced to thinking about the movement, which (surprisingly) has neocortex utilize a comparable amount of energy as when one uses the thinking to execute movement, and this energy consumption is ~ 20 times greater per neuron in the neocortex than it is per neuron in the cerebellum (at least across mammals, Herculano-Houzel 2010). This indicates that thinking/consciousness is a very energy expensive process.

(4) That consciousness depends on a functional neocortex and its synapses is an old story (Hebb 1968) given that anesthesia abolishes consciousness by disabling the synapses within critical association areas within the frontal and parietal lobes (Vigotsky et al. 2022). When electrical stimulation is delivered to the neocortex the signal travels trans-cortically by one synapse only, and not beyond (Tolias et al. 2005; Logothetis et al. 2010). What this means is that restricting the unit of consciousness up to one synapse away from a source neuron is supported by the neurophysiology of the brain, thereby reinforcing the idea that consciousness can be reduced to one neuron and its family of follower cells, but not by much more. Transmission beyond one synapse is prevented by GABAergic inhibition (Logothetis et al. 2010), which controls against eliciting epileptic discharges.

(5) The proposed model that one consciousness unit subserves one conscious event (or thought) will require further study:

First, methods will need to be devised to isolate and study a single unit and its connections to determine how consciousness at the unit level can be manipulated by altering the synaptic weights at the follower neurons which can also provide clues as to how much information (in bits per second) is represented by the activation of a generic consciousness unit once transmission time along the collaterals and synaptic times are deduced. Also, activation of a single unit should evoke one conscious fragment psychophysically (Tehovnik 2017) when electrical stimulation is delivered to an association area such as the parietal, temporal, or orbito-frontal cortex.

Second, the method by which transitions occur between two or more consciousness units needs to be ascertained to study James’s (1890) stream of consciousness quantitatively. This should resolve the age-old issue of how language and other serial processing behaviors are actualized at the neural level before being expressed at the muscles. The expression ‘I want to be a scientist’ is deducible by finding the neuron(s) encoding each word in neocortex and by determining the neural path by which these words are concatenated before being sent to the cerebellum and motor nuclei for full expression (Tehovnik, Hasanbegović, Chen 2024).

Third, some outstanding issues related to consciousness could be addressed from whether different consciousness units are recruited during consciousness versus unconsciousness in neocortex (Tononi et al. 2016), whether conscious memories are differentially consolidated in neocortex for the mentally impaired, and whether a mind-only preparation can be created to empirically assess René Descartes’s mind-body duality (see Footnotes 1-3).

Footnote 1: Tononi et al. (2016) have suggested that consciousness has a very low information transfer rate that is far below 40 bits per second. Their concept of information is intrinsic to parts of the brain supporting consciousness thereby excluding all unconscious processing, whereas Shannon’s formulation as used by information theorists is based on behavioral output which does not differentiate between conscious and unconscious states (Reed and Durlach 1998). Tononi et al. suggest that different cortical neurons may mediate conscious and unconscious states. A more parsimonious view would be to have the same neurons in neocortex mediate both processes (as retinal ganglion cells mediate both rod and cone vision) with the relatively more ‘active/connected’ neurons participating in the stream of consciousness (Tehovnik 2017). This is supported by the observation that hyperactive cortical activity that induces extreme tinnitus can hijack consciousness by disrupting both wakefulness as well as sleep (Axelsson and Sandh 1985; Han et al. 2009). Once overtraining is established for a task the number of cortical neurons involved in the execution of the task is diminished. We have estimated that during express saccades (a highly automated act) that the frontal lobes and the cerebellar cortex are excluded from the execution loop and that a direct V1-collicular pathway is sufficient to carry out the response (Schiller and Tehovnik 2015; Tehovnik, Hasabegović, Chen 2024).

When learning something new, the information transfer rate can be exceedingly low (e.g., ~ 0.002 bits per second to learn a language, Hosoda et al. 2013), which concurs with the surmise of Tononi and colleagues (2016) that consciousness can have a low information transfer rate. Learning a new task takes conscious effort even though much of the consolidation into long-term cortical storage occurs during slow-wave sleep (Wilson and McNaughton 1994; Ziyang, Sheth et al. 2017). After years of training in a discipline, the information transfer rate can surpass 40 bits per second as evidenced by an accomplished communicator or performer (Reed and Durlach 1998). Highly practiced athletes or musicians execute their ultimate performance automatically and with minimal consciousness. But in the end, it is the years of conscious dedication to a goal (as evidenced through learning and repetitive practice) that are responsible for achieving exceptionalism, as exemplified by Einstein, Kasparov, Pelé, and Bolt.

Footnote 2: The consolidation of memory during slow-wave sleep occurs at the neurons that participate in the learning process during wakefulness (Ziyang, Sheth et al. 2017), which makes perfect sense since why consolidate memories for neurons that did not participate in this process (which could conceivably happen to ‘patients’ like Donald Trump who suffers from a schizophrenic pathology and who as a consequence believes his own lies thereby making an insanity defense the only possibility for his lawyers).

Footnote 3: As long as the body has the propensity to move, the brain can generate a sufficient signal to move external devices (Tehovnik et al. 2013). This property was recently evaluated using monkeys that had received amputations of the forelimbs at an early age (Balasubramanian et al. 2017). It was found that unilateral forelimb-amputated monkeys could still generate an adequate cortical signal from the motor cortex to move a robotic arm, but signals from the motor cortex contralateral to the intact forelimb moved the device faster over the period of testing. It has been known for some time that once the proprioceptive inputs to the CNS are cut, monkeys fail to produce a signal in the motor cortex to move external devices (Wyler et al. 1978, 1979). Activation of the somatosensory cortex can permit volitional control of single cells in the motor cortex of a head-fixed mouse rewarded with water (Prsa et al. 2017). The challenge now is to determine whether such volitional control can be done by separating the brain from the body. This could be accomplished by replacing the reward delivery with electrical stimulation of the lateral hypothalamus (a reward center, Olds and Milner 1954) and by having all neural communication with the body severed and the brain maintained artificially. This would move us closer to assessing a mind-only preparation, a preparation that should fail to yield volitional control according to Birbaumer (2006).