The necessity for the observer to examine the system from a first-person frame of reference to trace the path of generation of inner sensations

Introduction

In contrast to other systems in the body, nervous system is unique in that several higher brain functions are first-person properties of the mind. These include the state of being conscious, the ability to perceive different sensory stimuli, and the ability to internally sense retrieved memories. Since only the owner of the nervous system has access to these, it has been difficult to obtain a mechanistic understanding of the generation of these first-person properties. Third-person observations have been carried out in detail at the biochemical, cellular, electrophysiological, systems, imaging and behavioral levels (Figure 1). Even though some of these findings from certain levels can find inter-connectable explanations, several others fail. This poses tremendous challenges in solving the system^1,2. Understanding the first-person property of perception can also influence other branches of science such as physics³. Is there any alternate method possible to understand how first-person internal sensations are being generated?

Figure 1. Frame difference between the first-person internal sensations and third-person views of the nervous system functions at different levels.

Current studies are third-person examinations and the results from them are used in correlational studies between them. Since most higher brain functions generate internal sensations, understanding the structure-function mechanism for inducing first-person internal sensations is required to understand the system.

The importance of first-person reporting for understanding the nervous system has been suggested by several investigators^4–6. First-person reports of inner sensations through motor activity such as behaviour and speech provide the third-person observers with surrogate markers about the content of internal sensations. This allows the observer to make certain judgements about the first-person inner sensations in a subject. The observer who is keen to understand the system examines the system closely at various levels. Again, the observer acquires different sets of surrogate makers of biochemical reactions, synaptic changes, cellular changes, neuronal activations, oscillating extracellular potentials, and signal changes from imaging studies. The observer then correlates these markers with the surrogate behavioural responses. These types of studies over several years have led to the implicit use of behavioral responses in lieu of the internal sensations of higher brain functions. Making conclusions about higher brain functions based on surrogate markers undermines the necessity to make scientific enquiries about the mechanism of formation of internal sensations.

Understanding how first-person inner sensations are generated has a direct impact on certain areas – memory disorders, mental disorders and usage of anesthetic agents. Some psychiatric disorders generate the first-person property of hallucinations with a compelling sense of reality. What conditions can autonomously generate a meaningful flow of sensory perception in the absence of any external sensory stimuli? Finding solutions for alleviating symptoms in these conditions requires an understanding of the normal mechanism of generation of first-person internal sensations. Anesthetic agents are being used to block the first-person property of consciousness and is being carried out in the absence of a mechanistic explanation for consciousness. This makes it difficult to understand how anaesthetic agents work and what might be leading to the neurodegenerative changes associated with their use⁷. Different attempts to build a framework for consciousness have been carried out^8–11.

How can the mechanism be discovered?

Both the previous methods of using first-person reports and correlational studies using different third-person observations^4–6 are short of providing a solution. A clear understanding of the mechanism of the generation of internal sensations should be able to guide the replication of this operation in an engineered system. The first attempt to build a framework for an empirical approach was initiated by Marvin Minsky¹². Since gradual changes from a simple cue stimulus to more complex ones generate corresponding internal sensations of retrieved memories at physiological time-scales, it indicates that the internal sensations are generated using unitary mechanisms. How can its mechanism be discovered? Reductionism can be used to carefully examine the system by keeping all the required constraints for making hypotheses regarding the smallest possible structure-function units for the induction of internal sensations. The operation of these units is expected to be part of a systems property. In this regard, views of emergent properties and reductionism can be seen as mutually inclusive. By using them in conjunction, the system may be approached to find a solution. The presence of a large number of diverse third-person observations at various levels indicates that the system has a unique solution. It is reasonable to expect that the solution is likely to be a simple one.

The barrier of accessing first-person inner sensations can be crossed by using theoretical approaches, which can be verified later. Selecting a higher brain function that is well-preserved across species and has been well-studied by third-person approaches is important to derive a probable mechanism. In this regard, learning and memory research offer a large number of observations at various levels – biochemical, cellular, electrophysiological, systems and behavioral. By setting up all the constraints, the theoretical approach can be carried out to examine the potential locations where learning-associated changes can take place between two stimuli. These changes are expected to be reactivated by one of the stimuli (cue stimulus) to induce the internal sensations of memory of the second item along with behavioral motor actions that would have occurred in the presence of the second stimulus alone^12,13. The constraints include operations occurring at specific frequency of oscillations of extracellular potentials and the ability to make signature changes for each associative learning event, changes that are capable of reversing (for forgetting) and that can be stabilized for different periods of time (for retaining memories for different periods of time).

At the identified specific locations and conditions, how does the cue stimulus induce a unit of internal sensation? Since the internal sensations consist of partial sensory features of the associatively learned item, how can it be sensed in the absence of the learned item? At the appropriate locations, the cue stimulus has to induce an operational mechanism that allows the system to sense the sensory qualia of the associatively learned item “from within.” This requires a search from the appropriate locations towards the sensory receptors that were activated by the learned item. This can be expected only through a retrograde extrapolation from those appropriate locations towards those sensory receptors that will identify the sensory inputs capable of activating those receptors. The net sensory content generated by the system in response to the cue stimulus constitutes the sensory content of the internal sensation for the associatively learned item. Here, the sensory qualia result from the approach of the system from a first-person frame of reference (Figure 2). An observer who traces along this path, from the identified locations to the sensory receptors, inevitably will be examining the system from a first-person frame of reference. The first-person internal sensations of other higher brain functions are also expected to share a similar principle and will require examination from a first-person frame of reference.

Figure 2. Flow chart showing the key feature of a mechanism for the generation of internal sensations at specific locations.

It is expected to be located at the points of convergence of associatively learned stimuli. The mechanism is derived by keeping all the constraints and is expected to induce first-person accessible internal sensations. To characterize the qualia of the internal sensation, it is necessary to extrapolate from the location of the mechanism (named as mechanism X) towards the sensory receptors. The observer has to follow the same route towards the sensory receptors to identify them and characterize the nature of sensory stimuli that are capable of activating them. This backward extrapolation from the location of the mechanism X towards the sensory receptors constitutes examination from a first-person frame of reference.

The gold standard

The gold standard for verifying the theoretically-derived operational mechanism requires its successful transfer to the engineered systems. Importance of developing engineered systems was explained previously¹⁵. These systems will provide the advantage of studying the nature of internal sensations by retrograde extrapolation towards the sensory receptors to find out the minimum sensory inputs capable of activating those receptors. The system can be built to provide readouts of the internal sensations and can examine whether they match with the behavioral motor outputs expected from the system (Figure 3). The nature of these sensory inputs will depend on the number of neuronal orders from the sensory receptors and the connections between them in a given engineered system. Using all the units of internal sensations, the algorithm required for computing them to obtain a partial sensory framework to identify the item whose memory is being retrieved can be obtained. In the case of lower order animals, the number, types and qualia of higher brain functions are expected to be limited since the number of sensory receptors and the locations of convergence of inputs are limited. By using regular feedback from computational studies to configure the algorithms for different higher brain functions, systems of different complexities can be built. At the advanced stages, the systems science will need to examine the systems properties from a holistic view, including its interaction with the surrounding environment and dynamic behavior through complex paths that are reinforced during certain operations. In addition, the systems science will be able to examine instabilities when the system crosses the “boundary conditions” that can mimic the disease process. Systems design, systems development, systems stability, systems analysis, systems dynamics, and systems viability will become necessary elements of this process.

Figure 3. Steps required to cross the barrier of frame difference between the first- and third-person sensible functions.

The schematic diagram shows a path for a scientific approach to replicate theoretically-feasible hypothesized mechanism in engineered systems. It is possible to set up readouts of the internal sensations generated in response to a cue stimulus from an engineered system, which can be used to fine-tune the systems properties. This approach is expected to lead to the development of artificially intelligent systems.

Major hurdles

The predictions made by the new approach can be verified, and followed by replication in engineered systems. Since the current studies are being carried out using third-person observations at various levels, any new approach that uses a change in the frame of reference in its methodology is expected to require time for the scientific community to examine the mechanism. This delay can be reduced by a) providing explanations for currently unexplainable observations in terms of the new mechanism, and b) by the experimental confirmation of the predictions made by the proposed mechanism. The arguments used in the theoretical derivation and supporting evidences are expected to motivate replication in engineered systems. Finally, concerns about “the singularity,” a threshold point above which engineered systems will become more intelligent than humans, will need appropriate actions and reassurance.

Conclusion

In order to trace the path through which the nervous system generates a first-person view, it is necessary to carry out an examination from a first-person frame of reference at some point during the investigational process. The virtual nature of the first-person internal sensations indicates that the steps towards successfully solving it will have similarities to the development of complex numbers in mathematics. The natural course of events that leads to the verification of the first-person properties in engineered systems and the development of artificial intelligence can be regarded as two sides of the same coin. These first-person studies will be unique to the field of neuroscience, compared to the studies of other organs in the body and should be brought to the mainstream investigational methods in the field. A discussion on this topic among neuroscientists, computational scientists, and engineers can spark many bright ideas.