Tag Archives: Bernstein

Consistency Of Task Outcome And The Degrees Of Freedom Problem-The Brain Is Potentially Not A Micro-Manager When Providing Solutions To Complex Problems

Part of the reason I enjoy cycling as my chosen sport now I am older is not just because it is beneficial from a health perspective, but because the apparent regularity of the rhythmical circular movement required for pedalling creates a sense of peace in me and paradoxically allows my mind to wander a bit away from its routine and usually work-focussed and life task orientated thoughts. I enjoy watching competitive darts, from the perspective of marvelling at how the folk participating in the competitions seem to so often hit the small area of the board they are aiming for with such precision, after fairly rapidly throwing their darts when it is their turn to do so. This week an old colleague and friend from University of Cape Town days, Dr Angus Hunter, published some interesting work on how the brain controls muscle activity during different experimental conditions, a field of which he is a world expert in, and it was great to read about his new research and innovative ideas as always. Some of the most fun times of my research career were spent in the laboratory working with Angus measuring muscle activity during movement related tasks, where one of our most challenging issues to deal with was the variability of the signal our testing devices recorded when measuring either the power output from, or electrical activity in, muscle fibres each time they contracted when a trial participant was asked to do the same task. A large part of the issue we had to solve then was whether this was signal ‘noise’ and an artefact of our testing procedures, or if it was part of the actual recruitment strategy the brain used to control the power output from the muscles. All of these got me thinking about motor control mechanisms, and how movement and activity is regulated in a way that gets tasks done in a seemingly smooth and co-ordinated way, often without us having to think about what we are doing, while when one measures individual muscle function it is actually very ‘noisy’ and variable, even during tasks which are performed with a high degree of accuracy, and how the brain either creates or ‘manages’ this variability and ‘noise’ to generate smooth and accurate rhythmical or target-focussed activity, as that which occurs when cycling and throwing darts respectively.

Some of the most interesting scientific work that I have ever read about was done by Nikolai Bernstein, a Russian neurophysiologist, who when working in the 1920’s at the somewhat euphemistically named Moscow Central Institute of Labour, examined motor control mechanisms during movement. As part of the communist government of the times centrally driven plans to improve worker productivity and output, Bernstein did research on manual labour tasks such as hammering and cutting, in order to try and understand how to optimise it. Using novel ‘cyclogram’ photography techniques, where multiple pictures were taken of a worker using a hammer or chisel to which a light source had been attached, he was able to produce the astonishing observation that each time the worker hit a nail or cut through metal, their arm movements were not identical each time they performed the action, and rather that there was a great degree of variability each time the similar action was performed, even though usually this variability in action produced an outcome which had a high degree of accuracy. He realized that each complete movement, such as moving the arm towards the target, is made up of a number of smaller movements of muscles around the shoulder, elbow and wrist joints, which together synergistically create the overall movement. Given how many muscles there are in the arm, working around three joints (and potentially more when one thinks of the finger joints and muscles controlling them), he suggested that were a very large number of potential combinations of muscle actions and joint positions that could be used for the same required action, and a different combination of these appeared to be ‘chosen’ by the brain each time it performed a repetitive task. From a motor control perspective, Bernstein deduced that this could potentially cause a problem for the brain, and whatever decision-making process decided on which movement pattern it would use to complete a task, given that it created a requirement for choosing a particular set of muscle synergies from a huge number of different options available, or in contrast not choosing all the other muscle synergistic options, each time the individual was required to perform a single task or continue performing a repetitive task. This would require a great amount of calculation and decision-making capacity on a repetitive basis by the brain / control processes, and he called this the motor redundancy, or degrees of freedom, problem.

Like a lot of work performed in the Stalin era in Russia, his fascinating work and observations did not become known to Western scientists until the 1960’s, when he published a text-book of his career in science, which was subsequently translated and taken forward by excellent contemporary movement control scientists like Mark Latash of the University of Pennsylvania State in the USA. Further studies have supported Bernstein’s earlier work, and it is astonishing how much variability there is in each movement trajectory of a complex action that is goal orientated. Mark has suggested that this is not a redundancy problem, but rather one of abundancy, with the multiple choices available being of benefit to the body of any individual performing repetitive tasks, potentially from a fatigue resistance and injury prevention perspective, which may occur if the same muscle fibres in the same muscle are used in the same way in a repetitive manner. Interestingly, when a person suffers a stroke or a traumatic limb injury, the quantity of movement variability appears to paradoxically reduce rather than increase after the stroke or injury, and this reduced variability of motor function is associated with a decrement in task performance accuracy and completion. Therefore, the high variability of movement patterns in healthy folk appears to paradoxically make task performance more accurate and not just more efficient.

How control processes choose a specific ‘pattern’ of muscle activity for a specific task is still not well known. A number of theories have been proposed (generally as a rule in science, the more theories there are about something, the more the likelihood there is that there is no clarity about it) with some quaint names, such as the equilibrium point hypothesis, which suggests that choice at the motor neuron level is controlled as part of the force-length relationship of the muscle; the uncontrolled manifold hypothesis, which suggests that the central nervous system focuses on the variables needed to control a task and ignores the rest (the uncontrolled manifold being those variables that do not affect task required activity); and the force control hypothesis, which suggests that the central nervous systems compares the required movement for the task against internal models, and then uses calculations and feedforward and feedback control mechanisms to direct activity against that set by the internal model; amongst others. All these are interesting and intellectually rigorous theories, but don’t tell us very much about exactly how the brain chooses a particular group of muscles to perform a task, and then subsequently a different group of muscles, which use a different flight trajectory, to perform the task again when it is repeated. It has been suggested that there are ‘synergistic sets’ of muscles which are chosen in their entirety for a single movement, and that the primitive reflexes or central pattern generators in the spinal cord may be involved. But the bottom line is that we just do not currently know exactly what control mechanism chooses a specific set of muscles to perform one movement of a repetitive task, why different muscles are chosen each time the same task is performed sequentially, or how this variable use of muscles for the same task is managed and controlled.

We have previously suggested that a number of other activities in the body beyond that of muscle control have similar redundancy (or abundancy) in how they are regulated, or at least in respect of which mechanisms are used to control them. For example, blood glucose concentrations can be controlled not only by changes in insulin concentrations, but also by that of glucagon, and can also be altered by changes in catecholamine (adrenaline or noradrenaline) or cortisol levels, and indeed by behavioural factors such as resisting the urge to eat. Each time blood glucose concentrations are measured, the concentrations of all these other regulatory hormones and chemicals will be different ratio-wise to each other, yet their particular synergistic levels at any one point in time maintains the level of blood glucose concentrations at homeostatically safe setpoint levels. The blood glucose level is maintained whatever the variability in the regulatory factor concentration ratios, and even though this variability in choice of control mechanisms similarly creates a potential for high computational load when managing blood glucose concentrations from a control perspective. Similarly, perception of mood state or emotions are thought to have redundancy in what factors ‘creates’ them. For example we can fairly accurately rate when we feel slightly, moderately or very fatigued, but underpinning the ‘feeling’ of fatigue at the physiological level can be changes in blood glucose, heart rate, ventilation rate, and a host of other metabolites and substrates in the body, each of which can be altered in a variable ratio way to make up the sensation of fatigue we rate as slightly, moderately or very high levels of fatigue. Furthermore, fatigue is a complex sensation made up of individual sensations such as breathlessness, pounding chest, sweating, pain, and occasionally confusion, dizziness, headache and pins and needles, amongst others, a combination of which can also be differently valenced to provide a similar general fatigue rating by whoever is perceiving the sensation of fatigue. To make it even more complex, the sensation of fatigue is related to inner voices which either rate the sensation of fatigue (the ‘I’ voice) or make a judgement on it related to social circumstances or family and environmental background (the ‘Me’ voice), and it is through the final combination of these that an individual finally rates their level of fatigue, which adds another level of redundancy, or abundancy, to the factors underpinning how the ‘gestalt’ sensation of fatigue is both created and perceived. There are therefore three potential ‘levels’ of redundancy / abundancy in the signals and factors which either individually or collectively make up the ‘gestalt’ sensation of fatigue, and a corresponding increased level of computational requirements potentially associated with its final genesis, and how this perceptual redundancy / abundancy is managed by the control mechanisms which generate them is still not well known.

In summary, therefore, the presence of variability during activities of daily living across a number of different body systems is not only ‘noise’ / artefacts of testing conditions which are challenges for us researchers to have to deal with, it also appears to be part of some very complex control mechanisms which must have some teleological benefit both for optimizing movement and activity, and ensuring the capacity to sustain it without fatigue or injury to the components of the mechanism which produces it. Each time I cycle on my bike and my legs move up and down to push the wheels forward, different muscles are being used in a different way during each rotation of the wheel. Each time a darts player throws a dart, different muscle synergies are used to paradoxically create the accuracy of their throw. There is real ‘noise’ that a researcher has to remove from their recorded traces after a testing session in a laboratory, such as that caused by the study participant sweating during the trial, which can affect electrophysiological signals, and there is always a degree of measurement error, and therefore some degree of ‘noise’ is present in the variability of the recorded output for any laboratory technique that measures human function. But, equally, Bernstein’s brilliant work and observations all those years ago helped us understand that variability is inherent in living systems, and after understanding this, each time I observe data, particularly that generated during electrophysiological work such as I have used for a number of experiments in my own research career, including electromyography (EMG), electroencephalography (EEG) or transcranial magnetic stimulation (TMS), which has low standard deviations in the results sections of published research articles, I do wonder at the validity of the data and whether it has been ‘paintbrushed’ by the researchers who describe it, as my old Russian neurophysiology research colleague Mikhail Lomarev used to describe it, when he or we thought data was ‘suspect’. The inherent variability in brain and motor control systems makes finding statistical significance in results generated using routine neurophysiological techniques more difficult. It also seems to create a huge increase in the requisite control-related calculations and planning for even a simple movement, though as Mark Latash suggested, the brain is likely to not be a micro-manager, but rather some effective parsing mechanism which can both generate and utilize a large number of synergistic movement patterns in a variable manner for any task, while not utilizing much decision making power using some sort of heuristic-based decision-making mechanism. Most importantly though, it fills one with a sense of awe at the ‘magic’ of our own body, and for the level of complexity involved in both its creation and operative management, when even a simple movement like striking an object with a hammer, or cutting a piece of metal, can be underpinned by such complex control mechanisms that our brains cannot currently comprehend or make sense of.

In a laboratory in the middle of Russia nearly a century ago, Nikolai Bernstein made some astonishing observations by doing exceptional research on basic motor control, while trying to increase the productivity of soviet-era industrial work. A century later we are still scratching our heads trying to understand what his findings mean from a motor control perspective. As I type these final sentences, I reflect on this, and wonder which synergistic composition of muscle activity in my fingers are responsible for creating the actions which lead to these words being generated, and realize that each time I do so, because of the concepts of variability, redundancy and abundancy, I will probably never use an identical muscle sequence when typing other ideas into words at another future point in time. But then again, I guess the words I will be writing in the future will also be different, and daily life, like motor control programs, will always vary, always change, even though the nail on the wall on which the picture hangs becomes a permanent ‘item’, as will this article become permanent when I hit the ‘send’ button to publish it. What is never to be seen again though are the traces in the ‘ether’ of the hammer blow which embedded the nail in the wall, and the exact movement of the individual muscles in the labourers arms and hands, and in my fingers as I typed which created these words. Like magic their variability was created, and like magic their pattern has dispersed, never to recur again in the same way or place, unless some brilliant modern day Bernstein can solve their magic and mystery, reproduce them in their original form using some as yet to be invented laboratory device, and publish them in a monograph. Let’s hope that if they do so, their great work does not languish unseen for forty years before being discovered by the rest of the world’s scientists, as was Bernstein’s wonderful observations of all those years ago!


Control of Movement And Action – Technically Challenging Conceptual Requirements And Exquisite Control Mechanisms Underpin Even Lifting Up Your Coffee Cup

During the Christmas break we stayed in Durban with my great old friend James Adrain, and each morning I would as usual wake around 5.00 am and make a cup of coffee and sit outside in his beautiful garden and reflect on life and its meaning before the rest of the team awoke and we set off on our daily morning bike-ride. One morning I accidentally bumped my empty coffee mug, and as it headed to the floor, my hand involuntarily reached out and grabbed it, saving it just before it hit the ground. During the holiday I also enjoyed watching a bit of sport on the TV in the afternoons to relax after the day’s festivities, and once briefly saw highlights of the World Darts Championship, which was on the go, and was struck by how the folk competing seemed with such ease, and with apparent similar arm movements when throwing each dart, to be able to hit almost exactly what they were aiming for, usually the triple twenty. When I got back home, I picked up from Twitter a fascinating article on movement control posted by one of Sport Sciences most pre-eminent biomechanics researchers, Dr Paul Glazier, written by a group of movement control scientists including Professor Mark Latash, who I regard as one of the foremost innovative thinkers in the field of the last few decades. All of these got me thinking about movement control, and what must be exquisite control mechanisms in the brain and body which allowed me to in an instant plan and enact a movement strategy which allowed me to grab the falling mug before it hit the ground, and allowed the Darts Championship competitors to guide their darts, using their arm muscles, with such accuracy to such a small target a fair distance away from them.

Due to the work over the last few centuries of a number of great movement control researchers, neurophysiologists, neuroscientists, biomechanists and anatomists, we know a fair bit about the anatomical structures which regulate movement in the different muscles of the body. In the brain, the motor cortex is the area where command outflow to the different muscles is directly activated, and one of the highlights of my research career was when I first used transcranial magnetic stimulation, working with my great friend and colleague Dr Bernhard Voller, where we able to make muscles in the arms and leg twitch by ‘firing’ magnetic impulses into the motor cortex region of the brain by holding an electromagnetic device over the scalp above this brain region. The ‘commands for action’ from the motor cortex travel to the individual muscles via motor nerves, using electrical impulses in which the command ‘code’ is supplied to the muscle by trains of impulses of varying frequency and duration. At the level of the individual muscles, the electrical impulses induce a series of biochemical events in and around the individual muscle fibres which cause them to contract in an ‘all or none’ way, and with the correct requested amount of force output from the muscle fibre which has been ‘ordered’ by the motor cortex in response to behavioural requirements initiated in brain areas ‘upstream’ from the motor cortex, such as one’s eyes picking up a falling cup and ‘ordering’ reactive motor commands to catch the cup. So while even though the pathway structures from the brain to the muscle fibres are more complex than I have described here – there are a whole host of ‘ancient’ motor pathways from ‘lower’ brainstem areas of the brain which also travel to the muscle or synapse with the outgoing motor pathways, whose functions appear to be redundant to the main motor pathways and may still exist as a relic from the days before our cortical ‘higher’ brain structures developed – we do know a fair bit about the individual motor control pathways, and how they structurally operate and how nerve impulses pass from the brain to the muscles of the body.

However, like everything in life, things are more complex than what is described above, as even a simple action like reaching for a cup, or throwing a dart, requires numerous different muscles to fire either synchronously and / or synergistically, and indeed just about every muscle in the body has to alter its firing pattern to allow the body to move, the arm to stretch out, the legs to stabilize the moving body, and the trunk to sway towards the falling cup in order to catch it. Furthermore, each muscle itself has thousands of different muscle fibres, all of which need to be controlled by an organized ‘pattern’ of firing to even the single whole muscle. This means that there needs to be a coordinated pattern of movement of a number of different muscles and the muscle fibres in each of them, and we still have no idea how the ‘plan’ or ‘map’ for each of these complex pattern of movement occurs, where it is stored in the brain (as what must be a complex algorithm of both spatial and temporal characteristics to recruit not only the correct muscles, but also the correct sequence of their firing from a timing perspective to allow co-ordinated movement), and how a specific plan is ‘chosen’ by the brain as the correct one from what must be thousands of other complex movement plans. To make things even more challenging, it has been shown that each time one performs a repetitive movement, such as throwing a dart, different synergies of muscles and arm movement actions are used each time one throws the dart, even if to the ‘naked’ eye it appears that the movement of the arm and fingers of the individual throwing the dart seems identical each time it is thrown.

Perhaps the scientist that has made the most progress in solving these hugely complex and still not well understood control process has been Nikolai Bernstein, a Russian scientist working out of Moscow between the 1920’s and 1960’s, and whose work was not well known outside of Russia because of the ‘Iron Curtain’ (and perhaps Western scientific arrogance) until a few decades ago, when research folk like Mark Latash (who I regard as the modern day equivalent of Bernstein both intellectually and academically) translated his work into English and published it as books and monographs. Bernstein was instructed in the 1920’s to study movement during manual labour in order to enhance worker productivity under the instruction of the communist leaders of Russia during that notorious epoch of state control of all aspects of life. Using cyclographic techniques (a type of cinematography) he filmed workers performing manual tasks such as hitting nails with hammers or using chisels, and came to two astonishing conclusions / developed two movement control theories which are astonishingly brilliant (actually he developed quite a few more than the two described here), and if he was alive and living in a Western country these would or should have surely lead to him getting a Nobel prize for his work. The first thing he realized was that all motor activity is based on ‘modelling of the future’. In other words, each significant motor act is a solution (or attempt at one) of a specific problem which needs physical action, whether hitting a nail with a hammer, or throwing a dart at a specific area of a dartboard, or catching a falling coffee cup. The act which is required, which in effect is the mechanism through which an organism is trying to achieve some behavioural requirement, is something which is not yet, but is ‘due to be brought about’. Bernstein suggested that the problem of motor control and action therefore is that all movement is the reflection or model of future requirements (somehow coded in the brain), and a vitally useful or significant action cannot either be programmed or accomplished if the brain has not created pre-requisite directives in the forms of ‘maps’ of the future requirements which are ‘lodged’ somewhere in the brain. So all movement is in response to ‘intent’, and for each ‘intent’ a map of motor movements which would solve this ‘intent’ is required, a concept which is hard enough to get one’s mind around understanding, let alone working out how the brain achieves this or how these ‘maps’ are stored and chosen.

The second of Bernstein’s great observations was what is known as motor redundancy (Mark Latash has recently suggested that redundancy is the wrong word, and it should have been known as motor abundancy), or the ‘inverse dynamics problem’ of movement. When looking at the movement of the workers hitting a nail with a hammer, he noticed that despite them always hitting the nail successfully, the trajectory of the hammer through the air was always different, despite the final outcome always being similar. He realized that each time the hammer was used, a different combination of arm motion ‘patterns’ was used to get the hammer from its initial start place to when it hit the nail. Further work showed that each different muscle in the arm was activated differently each time the hammer was guided through the air to the nail, and each joint moved differently for each hammer movement too. This was quite a mind-boggling observation, as it meant that each time the brain ‘instructed’ the muscles to fire in order to control the movement of the hammer, it chose a different ‘pattern’ or ‘map’ of coordinative muscle activation of the different muscles and joints in the arm holding the hammer for each hammer strike of the nail, and that for each planned movement therefore, thousands of different ‘patterns’ or ‘maps’ of coordinated muscle movement must be stored, or at least available to the brain, and a different one appears to be chosen each time the same repetitive action is performed. Bernstein therefore realized that there is a redundancy, or abundancy, of ‘choice’ of movement strategies available to the brain for each movement, let alone complex movement involving multiple body parts or limbs. From an intelligent control systems concept, this is difficult to get one’s head around, and how the ‘choice’ of ‘maps’ is made each time a person performs a movement is still a complete mystery to movement control researchers.

Interestingly, one would think that with training, one would reach a situation where there would be less motor variability, and a more uniform pattern of movement when performing a specific task. But, in contrast, the opposite appears to occur, and the variability of individual muscle and joint actions in each repetitive movement appears to maintain or even increase this variability with training, perhaps as a fatigue regulating mechanism to prevent the possibility of injury occurring from potentially over-using a preferentially recruited single muscle or muscle group. Furthermore, the opposite appears to happen after injury or illness, and after for example one suffers a stroke or a limb ligament or muscle tear, the pattern of movements ‘chosen’ by the brain, or available to be chosen, appears to be reduced, and similar movement patterns occur during repetitive muscle movement after such an injury, which would also be counter-intuitive in many ways, and is perhaps related to some loss of ‘choice’ function associated with injury or brain damage, rather than damage to the muscles per se, though more work is needed to understand this conceptually, let alone functionally.

So, therefore, the simple actions which make up most of our daily life, appear to be underpinned by movement control mechanisms of the most astonishing complexity, which we do not understand well (and I have not even mentioned the also complex afferent sensory components of the movement control process which adjust / correct non-ballistic movement). My reaction to the cup falling and me catching it was firstly a sense of pleasure that despite my advancing age and associated physical deterioration I still ‘had’ the capacity to respond in an instant and that perhaps the old physical ‘vehicle’ – namely my body – through which all my drives and dreams are operationalized / effected (as Freud nicely put it) still works relatively okay, at least when a ‘crisis’ occurs such as the cup falling. Secondly I felt the awe I have felt at many different times in my career as a systems control researcher at what a brilliant ‘instrument’ our brains and bodies as a combination are, and whatever or whoever ‘created’ us in this way made something special. The level of exquisite control pathways, the capacity for and of redundancy available to us for each movement, the intellectual capacity from just a movement control perspective our brain possesses (before we start talking of even more complex phenomena such as memory storage, emotional qualia, and the mechanisms underpinning conscious perception) are staggering to behold and be aware of. Equally, when one sees each darts player, or any athlete performing their task so well for our enjoyment and their success (whether darts players can be called ‘athletes’ is for another discussion perhaps), it is astonishing that all their practice has made their movement patterns potentially more rather than less variable, and that this variability, rather than creating ‘malfunction’, creates movement success and optimizes task outcome capacity and performance.

It is in those moments as I had when sitting in a beautiful garden in Durban in the early morning of a holiday period, reflecting on one’s success in catching a coffee cup, that creates a sense of wonder of the life we have and live, and what a thing of wonder our body is, with its many still mystical, complex, mostly concealed control processes and pathways regulating even our simple movements and daily tasks. In each movement we perform are concealed a prior need or desire, potentially countless maps of prospective plans for it, and millions of ways it can be actualized, from which our brain chooses one specific mechanism and process. There is surely magic in life not just all around but in us too, that us scientist folk battle so hard to try and understand, but which are to date still impenetrable in all their brilliance and beauty. So with a sigh, I stood up from the table, said goodbye to the beautiful garden and great friends in Durban, and the relaxing holidays, and returned to the laboratory at the start of the year to try and work it all out again, yet knowing that probably I will be back in the same place next year, reflecting on the same mysteries, with the same awe of what has been created in us, and surely still will no further to understanding, and will still be pondering, how to work it all out – though next year I will be sure to be a bit more careful where I place my finished coffee cup!

%d bloggers like this: