Physics of Biological Action and Perception

Mark L. Latash

• Publisher: Academic Press
• Year: 2019
• Print ISBN:
  9780128192849
• eBook ISBN:
  9780128192856
• Topic: Biological Sciences
• Subtopic: Neuroscience
• Index: Biological Sciences

Physics of Biological Action and Perception helps researchers interested in exploring biological motor control from a physics or alternative viewpoint perspective. The book introduces the idea of parametric control as a distinguishing feature of living systems. Sections cover how the CNS creates stable percepts based on fuzzy and continuously changing signals from numerous receptors and the variable processes related to ongoing actions. The author also develops the idea of control with referent coordinates to stability of salient variables in fields typically united under the label of "cognition." Examples of this include communication (how the gist of a message is preserved despite variability of phrases), thought processes (how one can solve a mental problem via different logical routes), and playing chess (how one selects an optimal move given a position on the board). The book is written for researchers, instructors, clinicians and other professionals in all the fields related to biological movement and perception. - Presents a unifying theory of motor control based on physics - Encompasses action, perception and cognition - Discusses referent coordinates, kinesthetic perception and stability of actions - Identifies the importance of the CNS over computational brain function

1

Introduction

Many years ago, a great scientist, Nikolai Bernstein, decided to write a book about dexterity (Bernstein, 1996). Although all people intuitively understand what dexterity is, to use this term in a productive way that allows performing experimental and theoretical scientific inquiry, one has to define it. Bernstein spent the whole first chapter of the book building a definition of dexterity. He had two basic requirements for such a definition. First, it had to be operational. This means that the definition could be used to identify, or—even better—quantify, dexterity and distinguish it from “non-dexterity.” Second, the definition had to match the preexistent intuitive understanding of this word as closely as possible.

Can one build a definition for living system using a similar strategy? Having such a definition seems crucial for providing a foundation for many established sciences such as biology, psychology, neuroscience, biomechanics, etc. Of course, one can find many definitions of living systems (or simply life) in books, dictionaries, and on the Internet. There is a problem, however. Most of these definitions do not define living systems but rather enumerate their features such as adaptation, growth, homeostasis, metabolism, reproduction, separation from the environment (e.g., by the skin or by the cellular membrane), and some others. I purposefully put this list of features in the alphabetical order not to create an impression that there are more important and less important features. Other definitions use vague terms such as self-organization, which themselves need a formal definition. For example, does a growing crystal self-organize? It seems to do this. But most people would not claim that a growing crystal is a living system. It self-organizes in too a predictable way to qualify as living. Still other definitions focus on a particular feature of living systems, for example, their seeming violation of the Second Law of Thermodynamics and reduction of entropy. Of course, this reduction is local and achieved at the expense of the external world.

Let us start building a definition for a living system with introspection: When we see a moving object, typically it is relatively easy to distinguish a living object from a non-living one. Inanimate objects move in predictable ways: If one knows or can guess salient parameters of an inanimate object, its initial state, and external forces, it is possible to predict its motion with high certainty. In contrast, living objects move in a much less predictable way. In particular, they commonly walk or crawl uphill, fly against the wind, and swim against the current. In other words, they are active. This word does not mean, of course, that living objects violate basic laws of nature. At least, we have no compelling evidence for such violations. Nikolai Bernstein appreciated the fact that activity was the primary distinguishing feature of biological objects. He spent the last years of his life trying to create a new science, physiology of activity (Bernstein, 1966). Important insights into the foundations of activity, as a defining feature of living systems, were also made by a great mathematician Israel Gelfand and a brilliant physicist Michael Tsetlin (Gelfand and Tsetlin, 1962, 1966).

Let me use a more intuitive example. Consider a living animal, e.g., a frog, and an exact replica of that animal—a toy frog. These two objects behave very differently even if all their parameters are matched perfectly given our current level of knowledge and technical sophistication. The toy is very predictable in its motion, whereas the frog obeys only the famous Harvard's Law: “Under the most rigorously controlled conditions of pressure, temperature, volume, humidity, and other variables, the organism will do as it well damn pleases” (Bloch, 2003).

There is another distinguishing, and commonly overlooked, feature of living systems: They show stability of important characteristics of behavior while other characteristics may be highly variable, even noisy. Here, the word behavior is used in a very general sense, from motor action to perception, information exchange, and cognitive processes. Consider the following few examples.

When a person performs an action multiple times, typically, the important action characteristics show relatively reproducible trajectories, while characteristics of elements contributing to the action are much more variable. This was demonstrated by Bernstein about 100 years ago in his, now famous, study of professional blacksmiths (Bernstein, 1930). He asked his subjects to perform one of their typical labor movements—hitting the chisel held in the left hand with the hammer moved by the right hand (they were right-handed)—multiple times and recorded the kinematics of the joints moving the hammer as well as of the hammer itself. Note that these were perfectly trained subjects for this action, given that they had performed it hundreds of times a day for years. Bernstein reported that the intertrial variability of the hammer trajectory was much smaller than the intertrial variability of individual joint trajectories. Clearly, this is not a trivial observation. The brain cannot send signals directly to the hammer; it can only send signals to muscles that cause joint rotations. So, how can it be that the signals to muscles produced highly variable joint trajectories while the hammer trajectory was much more consistent across trials? It took researchers over half a century to come up with a feasible explanation for this phenomenon.

Imagine now that you walk along an art gallery and move your head to look at the paintings. During this action, all sensory signals informing on the position of the head and body in space—visual, vestibular, and somatosensory—change. Nevertheless, you have an adequate, stable perception that the external world is stationary. This perception stays stable even if your motion along the gallery is not self-produced, e.g., if you sit in a wheelchair and another person pushes you along the gallery.

Consider another very simple example: Imagine that you press with the palm against a stop without moving the arm. When you vary the force, signals from all relevant peripheral sensory endings (receptors) change due to changes in the muscle geometry, tendon force, skin deformation, joint capsule tension, activity of certain spinal neurons (e.g., gamma motor neurons), etc. However, all these changes fail to violate the stable, veridical perception that the arm does not move.

Imagine now that you describe to a group of friends an important episode of your life multiple times. One does not have to run a formal experiment to agree that multiple accounts of the same episode will vary in the composition, structure of individual phrases, vocabulary, prosody, etc. However, assuming that you try to convey the important features of the episode in each of the accounts, gist of the story will remain stable despite the variable means of its delivery.

Finally, if you are asked to perform a typical everyday task multiple times, for example, to go to a grocery store a few blocks away and buy several items on the shopping list, more likely than not, you will use different trajectories from the home to the store and inside the store to accomplish the task. Some of the variability of the trajectories may be due to external factors such as streetlights, cars, pedestrians, other customers in the store, lines to the cashiers, etc. But even if the streets were empty and you were the only customer, likely your trajectories would differ spontaneously, due to unpredictable changes in your intrinsic states such as thoughts. Despite all this variability, most likely you will end up with the same (or not very variable) set of products in the shopping bag in front of your home entrance door.

What is so special about stability of the outcome in all these examples? And why doesn't the system ensure the same stable outcome by using the same, similarly stable, means (contributions of elements)? The first question is relatively easy to answer: Of course, unstable actions (movements, percepts, messages, and thoughts) are next to useless in most everyday situations because of the changing external conditions and intrinsic states of the living system itself. For example, to catch a prey or to avoid a predator, one better has a very stable, veridical picture of the environment with the prey/predator moving in it and a mechanism allowing stable, reliable actions in the environment despite possible unpredictable factors such as wind, stepping on a pebble, seeing an unexpected object, etc. The second question is much less obvious. Indeed, why is it advantageous being noisy at the level of elements? Further, I will try to offer an answer to this question and to merge the notions of activity and stability into a single coherent scheme that allows distinguishing behaviors of living systems and inanimate systems.

2

Definition of living system

Within this book, I accept the physical approach to living systems. This means that the ultimate goal of this line of inquiry is to discover laws of nature that define processes within living systems and interactions between living systems and their environments (which may include other living systems). This goal is different, for example, from trying to design and build an artificial device (a robot or a prosthesis) that is able to reproduce certain salient characteristics of certain behaviors of a living system. I assume that living systems emerged and developed as a result of the natural evolutionary process, not a divine intervention or some other act of creation, for example, by aliens. As a result, these systems are expected to obey laws of nature, which may be defined as compact representations of our experiences, typically in the form of equations. The scientific approach focused on laws of nature is commonly addressed as physics. Accepting the physical approach is equivalent to creating an area of natural science dealing with living systems, Physics of Living Systems.

This term is not new. In fact, the author graduated many years ago in a department called “Physics of Living Systems” from the Moscow Institute of Physics and Technology, also known informally as Fiztekh. As I understand now, the name of that department had little to do with reality—physics of living systems did not exist at the time and hardly exists now—but rather represented a promise that maybe, at some time in future, alumni of that program would help to create this field of science.

The purpose of classical physics has been to unite our experiences into a relatively small set of basic laws. These laws are typically formulated as equations that link salient variables, which describe states of the systems of interest, and parameters. One may say that the variables are constrained by the laws of nature while parameters are not. For example, consider arguably the most famous law of nature, Newton's Second Law, F = m a. This equation links force vector (F) acting on an object and acceleration (a) of the object. These two variables are constrained by the law: If one of them changes, the other has to change as well. The parameter m (mass) describes specific objects and is not constrained by this law: The law is equally applicable to objects with large and small mass, and mass does not have to change with changes in force or acceleration.

Laws of classical physics are applicable to particular types of objects. For example, the mentioned Newton's Second Law is applicable to objects with inertia. The well-known Hooke's Law, ΔF = –kΔx, is applicable to certain classes of deformable objects that resist deformation and accumulate potential energy, commonly addressed as springs. In this equation, deformation of an object (Δx) is linked to a change in force (ΔF) between this object and the environment with the help of a parameter called stiffness (k > 0). Application of this equation to different classes of objects that do not deform but move (for example, animal joints and limbs) has led to much confusion in the field of biomechanics (reviewed in Latash and Zatsiorsky, 1993, 2016), including reports of negative stiffness, which makes no sense in classical mechanics.

Living systems are not expected to violate Fundamental Physical Laws (under this term I understand all the laws found in physics textbooks). These laws impose constraints on biological objects, but these objects are not driven by those laws. In his recent book, Misha Gromov (2018) states that compatibility with laws of physics (the author implied Fundamental Physical Laws) is only a miniscule part of the foundation for biology. Gromov cites Erwin Schrödinger, who in his classic “What Is Life?” (Schrödinger, 1944, 2012) suggested that living objects were likely to obey unknown to us laws of physics, which, after they have been discovered, would form an inherent part of physics in general. Along similar lines, Israel Gelfand stated that the problem was not in applying known to us mathematics to biology but in developing new biological chapters of mathematics (quoted in Latash, 2008).

It is commonly assumed in classical physics that, during typical times of observation, parameters of systems of interest remain unchanged or change at much slower rates compared with law-constrained variables. I am going to suggest that living systems differ from inanimate ones in a major way: Living systems induce changes in their states by changing parameters of relevant biological laws of nature. These are going to be addressed as Biological Physical Laws in contrast to the Fundamental Physical Laws, common for all objects, living, and inanimate. Here comes a simple, brief definition, which is a step toward development of Physics of Living Systems:

A living system is a system able to (1) unite Fundamental Physical Laws i...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Preface
• The preamble
• 1 Introduction
• 2 Definition of living system
• 3 Redundancy and abundance
• 4 Stability in abundant systems
• 5 Control with referent coordinates
• 6 Kinesthetic perception
• 7 Stability of Gist
• 8 Problems and conjectures
• References
• Index