The necessity of studying higher brain functions from a first-person frame of reference

Introduction

Attempts to interconnect third-person findings obtained at different levels such as biochemical, cellular, electrophysiological, systems, imaging and behavioral studies by different fields of neuroscience remains a challenge^1,2. In contrast to other systems in the body, nervous system functions are unique in that all the higher brain functions are first-person properties of the mind. These functions include the state of being conscious, the ability to perceive sensations, the ability to internally sense retrieved memories and the ability to generate thoughts (Figure 1). Only the owner of the nervous system has access to these functions, making them purely first-person internal sensations³. First-person reports of these sensations through motor activity such as behaviour and speech provide surrogate markers to the third-person observers. Currently, clinical evaluation of neurological and psychiatric diseases is based on assessing the first-person reporting and third-person observed findings. This severely limits our understanding of a) internal sensations in non-responsive patients, b) defects in the mechanism of formation of the internal sensation of memory, and c) the compelling sense of reality of hallucinations in psychiatric disorders. The surrogate markers for assessing the first-person properties may not represent the true nature of the internal sensations. This is because a) behaviors may be altered due to changes in the circuitry, or b) some species of animals may voluntarily hide the truth and exhibit behaviors misrepresenting them. Similarly, almost all the current approaches use third-person observed findings at various levels⁴ in correlational studies with surrogate makers of biochemical changes, neuronal activations, oscillating potentials, signal changes in imaging studies, and behavioural responses to connect with the first-person properties. Another important area of study is to understand consciousness⁵. An entire branch of medicine that deals with blocking this first-person property of consciousness – anesthesiology – requires knowledge of its mechanism due to several reports of neurodegenerative diseases associated with anesthetics⁶.

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

Third-person observed features include neuronal firing, electrophysiological changes and surrogate markers of internal sensations such as behavioral motor activities and language. It has not been possible to interconnect between different third-person findings. Hypothesizing the mechanism of first-person internal sensations of different higher brain functions capable of interconnecting various third-person observed findings is an optimistic step towards verifying the first-person view.

Current third-person studies at various levels assume that first-person internal sensations are emergent properties of the system. Emergence can be adopted as a framework to study properties that cannot be explained using the third person-observed features of the system. However, reductionism can be used to carefully examine the factors upon which the emergent properties are dependent and hypothesize the smallest possible structure-function unit from which internal sensations can be induced. Therefore, views of emergent properties and reductionism can be seen as mutually inclusive. By using them in conjunction, first-person internal sensations can be approached.

Converting first-person sensations to third-person features

The first-person features of the higher brain functions cannot be studied in biological systems due to access issues. Recent research work is attempting to overcome this barrier by approaching higher brain functions from a first-person frame of reference, examining the locations and mechanisms that can lead to the formation of the basic units of internal sensations^7,8. This drastically different approach is based on the view that the gold-standard test of understanding the formation of first-person internal sensations is to replicate the mechanism in engineered systems. This approach is being carried out in three stages. The first step is the theoretical derivation of the basic functional units of the system at the correct level that is also connected to the motor system, which can explain all the higher brain functions along with behavioral motor activity. It is found that a) locations from which memories can be retrieved gradually shift from the hippocampus to the cortices over several years, and b) patients recover completely after suffering from small strokes at certain locations of the brain. These suggest that the basic structure-function units are spatially definable and transferable, and that emergent functions can be integrated from multiple locations. Since a large number of functions and loss-of-function states for the system are being studied by different faculties of brain sciences, the solution capable of explaining both the first- and third-person properties is likely a unique one. In other words, there is only one solution. Therefore, theoretical work to hypothesize structure-function units is the first major step. The second step is to carry out computational studies to examine the nature of the algorithms for different modules of functions that can result in expected qualities for the generated internal sensations. The third step is to build engineered systems that can provide readouts of the formed internal sensations based on the rules by which they are built (Figure 2). It is likely to require combining the second and third steps.

Figure 2. Steps required to cross the barrier of frame difference between the first- and third-person functions.

Diagram shows a path for the first-person scientific approaches for replicating theoretically feasible hypothesized mechanisms in engineered systems. The readouts obtained from these systems can be used by third-person experimenters to fine-tune the internal sensations to match with both the expected internal sensations and behavioral motor activity. An inevitable end-product of this approach is the development of artificially intelligent systems.

Focal points of emergence

Of all the third-person sensed findings, neuronal firing (also known as somatic spikes) can be easily observed, induced and measured. New tools to make the firing of neurons visible provide an advantage in examining the nervous system from the third-person frame of reference. Somatic spikes are one of the different kinds of spikes observed along the neuronal processes. Others are dendritic spikes and axonal spikes. Most importantly, the potentials originated by the dendritic spike at farther locations degrade as they arrive at the neuronal soma. When examined from the third-person frame of reference, it can be seen that a very large number of excitatory postsynaptic potentials (EPSPs) are not being used efficiently to justify their evolutionary preservation. For example, EPSPs during sub- and supra-threshold activations of a neuron are not contributing to any function. Contribution of potentials from synaptic events that occurs remote from the neuronal soma towards neuronal firing is minimal⁹. Is there a different view possible for their functional attributes? The evolutionarily well-preserved occurrence of all the synaptic potentials, in surplus to what is required for the observed neuronal firing, prompts some important questions. What functional significance can they impart when examined from a first-person frame of reference? For such investigations, the most important question is “At what focal points in the nervous system do the units of internal sensations emerge?” These approaches are expected to ultimately guide the discovery of units for the generation of internal sensations.

The gold standard

The gold standard for understanding the operational units requires transferring the theoretically-derived operational mechanism in engineered systems. Since internal sensations are virtual in nature, the conversion of the formed internal sensations into third-person observable outputs is indispensable to understanding the operation. In this context, the main limiting step is the theoretical derivation of the basic operational unit, at the correct level, that can explain all the nervous system functions from both first- and third-person frames of reference. This is followed by replication of the mechanism in engineered systems similar to that which are proposed¹⁰. Studies of first-person systems will deal with optimizing the properties of the components of the engineered system by seeking specific experimental results from biological systems. These studies will require regular feedback from computational studies to solve optimization problems. Finally, the studies are expected to arrive at the algorithms that can provide the desired outputs. At the advanced stages, the systems science will examine the systems properties from a holistic view, including its interaction with 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.

Major hurdles

There are two major obstacles in exploring the first-person properties of the nervous system. In my opinion they cannot be separated from the very challenge of discovering the first-person properties. The first one is reaching a consensus among researchers of one faculty of science for funding projects that involve significant changes in the research approach. If a logically-fitting experimental approach is available, then practical difficulties in conducting it should not deter us from undertaking it. Since the mechanism of the nervous system functions has not discovered yet, it is understandable that a novel approach is required. Changing the frame of reference from which to examine the higher brain functions suits such an anticipation. In my opinion, first-person approach should be brought into the main stream investigational methods in neuroscience. In fact, first-person studies should distinguish neuroscience from the studies of other organs in the body.

The second challenge is to maintain a certain level of confidence that we can discover the mechanism of formation of first-person properties. The necessity of discovering the mechanisms for the disorders of the mind should take precedence over the fear of discovering the operations of the mind. There is a growing concern about ‘the Singularity’, a threshold point above which engineered systems will become more intelligent than humans. The fear comes from the thinking that artificially intelligent machines may take over the human race. Building regulatory bodies and having strategies in place to prevent the development of these foreseeable effects should be simultaneously carried out along with the development of methods for exploring the first-person properties.

In conclusion

A reasonable early expectation from first-person studies is the development of experiments that can provide third-person-sensible outputs at key stages, so that data collection and exploration of further work can become possible. Comparative physiology of the mechanisms of formation of first-person properties using different model nervous system circuitries will be part of this approach. The first-person studies will also aim to identify the focal points at which the mechanism is disrupted in neurological and psychiatric disorders. By taking advantage of the information arriving from first-person studies, we will be able to design methods to prevent and treat several neurological and psychiatric diseases.

The nervous systems of even very low-level species produce intentionality to carry out survival and reproductive instincts, indicating that an evolutionarily highly conserved mechanism is shared among all species of animals. The presence of nearly ten million existing and predicted animal species on earth¹¹ provides a great deal of confidence to successfully simulate the mechanism in engineered systems that mimic one of them. Such intelligent systems are of paramount importance to help aging populations with the care they need, design strategies to feed the hungry, cure diseases, alleviate human suffering and provide methods to prevent climate change, to name a few. My opinion is that the steps towards finding solution to the virtual nature of the first-person properties will have similarities to the development of complex numbers in mathematics. Therefore, first-person studies will fall into the realm of a completely independent new branch of basic science. The conclusion of this opinion article is that a first-person approach to understand the brain and the natural course of events that will lead to the development of artificial intelligence are two sides of the same coin. A discussion on this topic among neuroscientists, computational scientists, and engineers can spark many bright ideas.