I first encountered the ideas of Nikolai Bernstein while studying psychology at university. Our programme placed considerable emphasis on general psychology, theories of higher nervous activity, the central nervous system and human psychophysiology. At the time, these subjects felt like part of a solid psychological foundation: important, certainly, but perhaps somewhat academic. 

Later, when I began working in the corporate world as a learning and development specialist, I increasingly noticed a strange gap. Corporate training for adults is often designed and delivered by people who are highly competent in course development, presentations, tests, competency matrices and training documentation, but who may have little grounding in the basic questions of psychology, psychophysiology, skill formation and human action. 

As a result, the way people are treated as the subject of training can become surprisingly mechanistic. 

It is as if the human being is seen as a system into which an instruction must be loaded. If someone has attended a course, seen a presentation, repeated a procedure and passed a test, then they are considered trained. If they make a mistake afterwards, the conclusion is often that they were inattentive, insufficiently disciplined or failed to follow the procedure. 

But people do not work like machines receiving commands and reproducing programmes. 

This is particularly visible in the oil and gas industry, where operators, technicians, mechanics, electricians, instrumentation specialists, supervisors and HSE professionals work in real, changing and often uncertain environments. In such conditions, it is not enough simply to remember a procedure. A person must understand the task, read the conditions, notice feedback, recognise weak signals, adjust their actions, make decisions under pressure and retain meaningful control over what is happening. 

That is why I believe it is worth returning to Bernstein today — not merely as a historical figure from Soviet physiology, but as a thinker who may prove unexpectedly useful to the modern corporate world. 

His ideas can help learning specialists, heads of training centres, HSE managers, operations leaders and technical instructors revisit a deceptively simple question: what is a skill, really? 

A scientist who studied not movement, but action 

Nikolai Aleksandrovich Bernstein was a Soviet physiologist and psychophysiologist, and one of the most important researchers of movement, coordination, dexterity and skill formation. He was born in Moscow in 1896 and died there in 1966. His scientific life unfolded during a profoundly difficult period: revolution, civil war, the early Soviet project of organising labour, the Stalinist era, post-war science and the gradual rehabilitation of lines of thought that had previously developed under intense ideological pressure. 

Bernstein trained as a medical doctor, served as a physician, worked in a psychiatric clinic and later entered an environment in which the study of human beings was linked not to abstract theory, but to practical questions of labour, movement, sport, rehabilitation and effective action. At the Central Institute of Labour, he studied the movements of workers engaged in manual labour, including the use of a chisel and hammer. Later, he studied walking, running, sporting movements, the movements of people with brain injuries and various forms of complex coordination. 

This context matters. Bernstein did not arrive at his ideas through abstract philosophy. He arrived at them by observing real human action: labour movements, walking, sporting technique, recovery after injury and bodily coordination under complex conditions. He studied the human being not as a passive object responding to stimuli, but as an active system that constructs action in pursuit of a goal. 

This is why his approach became known as the physiology of activity. 

In Bernstein’s time, Soviet science was subject to strong ideological control. Physiology and psychology, especially during the Stalinist period, were not neutral academic territories. They were shaped by dominant doctrines, above all the Pavlovian reflex paradigm. For scientists such as Bernstein, it was not easy to develop a more complex understanding of the human being — not as a bundle of reflexes, but as an active subject of action. 

It would be too simplistic to say that Bernstein worked on labour movements, sport and biomechanics only because these fields were “safer”. But applied fields did give him a language and a space in which to explore a much deeper question: how does a person construct meaningful action in the real world? 

At first glance, he studied movement. In reality, he studied action. 

And that has direct relevance for corporate training. 

Skill is not a memorised procedure 

In the corporate world, the word “skill” is used so frequently that it almost ceases to require explanation. We speak about technical skills, operational skills, safety skills, troubleshooting, maintenance, decision-making, communication, human factors and human performance. We build competency matrices, run training, issue certificates, assess operators, train supervisors and investigate mistakes. 

But if we look at all this through Bernstein’s lens, one thing becomes clear: a skill is not simply knowledge of a procedure, nor is it the mechanical repetition of a correct sequence. 

A skill is the ability to reconstruct the right action, again and again, under real conditions. 

This is especially important in the oil and gas industry, where the cost of error can be very high. Operator error, maintenance error, incorrect isolation, poor shift handover, a formalistic toolbox talk, weak risk assessment or delayed recognition of an abnormal situation can lead not merely to reduced efficiency, but to serious safety consequences. 

A person does not act like a machine that receives a command and reproduces a fixed programme. Even when the external result appears the same, the internal construction of the action is slightly different each time. Body position changes. Equipment condition changes. Context changes. Attention changes. Fatigue changes. Information quality changes. Communication with others changes. A good professional does not simply repeat an action. They reconstruct it, constantly checking it against the task, the conditions and the feedback available. 

Bernstein described this, in relation to movement, as “repetition without repetition”. From the outside, the action may look like repetition. Internally, however, it is a new solution to a task each time. 

For corporate training, this idea is hugely important. If we give workers the same training scenario again and again, they may become very good at that particular scenario. But that does not necessarily mean they are ready for real work. 

An operator who has practised one version of a high-pressure alarm response ten times may struggle if, in reality, the high-pressure alarm is not the cause but the consequence of another problem. A technician who habitually links pump vibration to a bearing issue may miss cavitation, misalignment, foundation problems or a change in process conditions. An HSE professional may formally review a permit to work without noticing that the actual conditions on site have already changed. 

This is why strong training should not be built around identical repetition, but around variable task-solving. 

First, people need a foundation: the principle, the procedure, the standard example. But then the training must become more demanding: similar but different scenarios; incomplete information; misleading cues; time pressure; conflicting priorities; changing conditions; and situations where the right answer is to stop and ask for support. Only then do we develop not the memory of an answer, but the ability to think and act. 

Action always begins with a task 

One of Bernstein’s key ideas is the concept of the motor task. Movement does not begin as a list of commands to the muscles. It begins as a task. A person does not simply “move an arm”. They open a valve, tighten a bolt, press an emergency stop, connect a cable, check an instrument, maintain balance or pass a tool. The movement has meaning because it is directed towards a result. 

This may sound obvious, but it has serious implications for training. 

Many technical programmes break work down into a sequence of steps. This is necessary, but not sufficient. If a trainee learns the steps without understanding the task, the result is fragile competence. The person knows what to do in a standard scenario, but becomes uncertain when conditions change. 

A valve is not opened because it is “step 4”. It is opened to establish flow, isolate, depressurise, redirect, protect, drain, test or restore a system. 

Lockout/tagout is not a paperwork ritual. It is the control of hazardous energy. 

A permit to work is not an administrative form. It is a way of aligning people, hazards, controls, responsibilities and changing site conditions. 

A pre-start-up check is not a checklist exercise. It is the final verification that a system is ready for safe operation. 

When training focuses only on visible steps, it can produce compliance without understanding. When training focuses on the task, it develops professional judgement. 

The same logic must be applied to cognitive skills. 

An operator does not simply “know process parameters”. They must understand what is happening to the plant, whether a deviation is normal or dangerous, what data supports a hypothesis, what actions are permissible and when the situation must be escalated. 

A maintenance specialist does not simply “know the pump”. They must understand why the pump is behaving differently: whether the issue is a bearing, cavitation, misalignment, lubrication, vibration from nearby equipment, operating mode, a false instrument reading or a misinterpretation of the symptom. 

A supervisor does not simply conduct a toolbox talk. They must understand whether the team is ready for the job, whether people have heard the key risks, whether there is fatigue or haste, whether responsibilities are clear and which information must be rechecked before work starts. 

A professional skill is the ability to solve a professional task in specific conditions. Sometimes the task is physical. Sometimes it is cognitive. Most often, it is both. In real work, an operator, technician or supervisor almost never performs only a physical or only a mental action. They perceive, think, check, move, communicate, correct and think again. 

Cognitive skills: action in thought 

Bernstein studied movement, coordination and motor skill. It would therefore be wrong to claim that he directly created a full theory of corporate decision-making or troubleshooting. But his logic — task, feedback, correction, variability, automation and error as learning material — can be applied very productively to cognitive skills. 

A cognitive skill can be defined as the ability to recognise a situation, build a working mental model, choose a course of action, test it against feedback and adjust the decision when conditions change. 

This applies to diagnosis, troubleshooting, risk assessment, shift handover, incident investigation, abnormal situation management and decision-making under uncertainty. 

A good operator does not simply react to the loudest alarm. They ask: what changed first? Which parameters are connected? What is the cause and what is the consequence? What data supports my hypothesis? What must I check before intervening? When should I call the field operator or supervisor? 

A good technician does not jump straight to repair. They build a diagnostic map: what changed, when it began, which parameters are connected, what has already been checked, which test will provide the most information with the least risk, which actions must not be performed before isolation, and where the stop point is. 

A real risk assessment is not the filling in of a form. It is cognitive work: constructing a model of a future action. What are we about to do? What could go wrong? What energy is present? Who is involved? What conditions might change? Which controls are actually working? How will we know that the job is no longer safe? 

A shift handover is not simply a list of events. It is the transfer of a mental model. What is happening now? Which deviations are developing? What actions have already been taken? What needs to be watched? Which risks may appear later? Which assumptions remain untested? 

If a shift handover transfers facts but not understanding, the next shift may begin work with the wrong model of the situation. 

Without feedback, there is no skill 

For Bernstein, action is not built through a simple chain of “command — execution”. It is built through continuous sensory corrections. A person acts, receives information about the result, compares it with the goal and adjusts the action. 

The same is true in professional work, except that feedback is not only bodily; it is also informational. 

In a control room, feedback comes from alarms, trends, process parameters and messages from the field team. In maintenance, it comes from sound, vibration, temperature, smell, leakage, instrument readings and equipment response. In electrical work, it comes from confirmation of isolation, tester readings, panel status and the match between drawings and the actual installation. In supervision, it comes from team questions, signs of fatigue, communication quality and changes on site. In risk assessment, it comes from the gap between the document and the real job. 

If feedback is absent, delayed, overwhelming or misleading, the probability of error rises. 

That is why human error is rarely just “a bad action by a bad worker”. Very often, it is the result of someone constructing action under poor conditions: incomplete information, weak feedback, an awkward procedure, a poor interface, fatigue, time pressure or insufficient understanding of the task. 

This leads to an important conclusion for HSE: human error should not be reduced to individual blame. We need to understand how the action became possible under those conditions. Why did the decision seem reasonable to the person at the time? Which signals did they see? Which did they miss? Which hypothesis became dominant? Why did they not revise it when new information appeared? What in the system pushed them towards that action? 

In this view, error is not only a violation. It is diagnostic material. 

If a trainee makes a diagnostic mistake, it is not enough to say: “That is the wrong answer.” We need to understand why the wrong answer became logical to them. Did they focus on the first visible cue? Did they ignore a weak signal? Did they overvalue one source of information? Did they fail to test an alternative hypothesis? Did they trust past experience too much? Did they rush because of time pressure? Did they use the checklist as a formality rather than as a way of testing reality? 

For learning, this is fundamental. A good instructor does not merely correct an error. A good instructor studies it. 

The invisible steps of mastery 

Bernstein also helps us rethink the development of mastery. In his description, skill formation passes through several stages. First, the person forms an overall idea of the action and performs it in broad outline. Then comes deeper learning: the action is refined, errors are reduced and individual elements begin to become skill. Finally, at the level of mastery, the action becomes more stable, economical and reliable, and adapts to different conditions of performance. 

This model is relevant not only to movement. It helps us understand almost any complex professional capability: operating a plant, diagnosing equipment, assessing risk, conducting a briefing, handing over a shift, making decisions or learning a foreign language. 

In the early stages, progress is usually easy to see. A person cannot do something; then they begin to understand it; then they can do it with guidance; then they can do it independently in simple conditions. These steps are large and visible. 

But the higher the level, the finer the differences become. 

The difference between a novice and someone who can “do it in broad outline” is enormous. The difference between a good professional, a very good professional and a master may be almost invisible to an outside observer. It does not show itself in the number of procedural steps memorised, but in the quality of action: the master recognises meaningful cues faster, distinguishes cause from consequence more accurately, notices weak signals earlier, understands sooner when a situation is no longer standard, remains calmer under pressure and corrects a mistaken hypothesis more quickly. 

This is why advanced skills require a more refined scale of assessment. 

Language learning illustrates the point well. The A1, A2, B1, B2, C1 and C2 levels are useful. But within B2 or C1 there may be many different states of ability. One person may write business emails confidently but struggle with fast speech in meetings. Another may speak well but make mistakes in technical writing. A third may understand general speech but struggle in a professional discussion about maintenance or safety. Formally, all may be “B2”, but for a corporate task they represent different levels of readiness. 

The same is true in oil and gas. The statement “the operator is competent” may be too crude. Competent in what exactly? Standard operation? Start-up and shutdown? Recognising deviations? Managing alarm flooding? Communicating with the field team? Diagnosing an abnormal situation? Acting with incomplete information? Remaining in control under pressure? 

At a basic level, it may be enough to say “knows the procedure” or “can perform under supervision”. At a higher level, that is not enough. We need to distinguish between knowledge, initial ability, independent performance, stable performance, adaptive performance, expert performance and mastery under complex conditions. 

There is an important paradox here: the higher the level of mastery, the smaller the visible step of progress — but the greater its professional value. 

Two operators may both know the procedure, pass the test and complete the standard simulator exercise. But one may notice a developing deviation ten minutes earlier, while the other notices it only after an obvious alarm. One may realise that an instrument reading does not match field conditions; the other may accept it as reliable. One may stop and ask for verification; the other may continue along the familiar path. Formally, both may be “competent”. In practice, the difference is enormous. 

Technology as an external feedback loop 

This is where modern technology can become a new external feedback loop. 

If Bernstein showed that skill is formed through action, feedback and correction, then wearable sensors, video analytics, voice assistants, augmented reality, simulators and digital learning platforms may help make this process more visible. 

In the past, an instructor could observe only part of the action: whether a step was completed, how long it took and whether an obvious error occurred. But many elements of skill remained hidden: micro-movements, delays in attention, signs of overload, the quality of information search, the sequence of decisions, the moment an error began, changes in confidence and response to feedback. 

Wearable sensors can capture body position, pace of movement, stability, movement around a site, proximity to hazardous areas and possible signs of fatigue or overload. Video analytics can help assess action sequence, use of personal protective equipment, safe distance and worker position relative to equipment. Voice assistants and communication analysis can reveal the quality of shift handover, clarity of commands, confirmation of understanding, communication gaps and delays in decision-making. Augmented reality can provide prompts in the worker’s field of view, warn of hidden hazards and help compare the actual state of equipment with the procedure or drawing. Simulators can create variable scenarios in which not only procedural knowledge is tested, but also situation recognition, hypothesis-building and decision correction. Digital platforms can show the trajectory of skill development: which errors repeat, which scenarios cause difficulty, where progress is being made and where the learner has reached a plateau. 

But it is easy to get this wrong. Technology must not turn training into total surveillance. Nor should it create the illusion that mastery can be reduced to a single number. 

The right question is not how to control every worker. The right question is how to give the worker, the instructor and the organisation more precise feedback on skill formation. 

For physical skills, technology may help us see how a person approaches equipment, how they use a tool, whether they make unnecessary movements, whether they apply excessive force, whether they maintain a stable body position, whether they miss critical checks and how performance changes under fatigue or time pressure. 

For cognitive skills, the task is more difficult: thought cannot be measured directly like the movement of a hand. But traces of thought can be observed — in information selection, speech, action sequence and response to changing conditions. In operator training, we can analyse which parameters a person looks at first, how quickly they notice a deviation, which signals they treat as important, whether they test alternative hypotheses and when they decide to escalate. In technician training, we can examine where diagnosis begins, which checks are chosen, whether the person jumps straight to repair and whether they stop under uncertainty. In supervisor training, we can observe which questions are asked before work starts, whether team understanding is checked, whether changing conditions are noticed and whether space is created for stop-work authority. 

Yet technology should not make the final judgement on competence without professional interpretation. It can show data, identify recurring patterns, indicate where a difficulty arose and help a worker see their own progress. But the decision about competence must remain a professional judgement grounded in context. 

Otherwise, measurement replaces understanding. 

Not everything important is easy to measure. Not everything measurable is truly important. Not every deviation is an error. Not every delay is a weakness. Not all variability is bad. Not all standardisation is useful. From Bernstein’s perspective, variability is sometimes not the enemy of skill, but evidence of adaptation to real conditions. 

Poor use of technology says: “The algorithm decided the worker acted incorrectly.” Good use says: “The data shows that in this type of scenario the worker notices the key signal later. Let us understand why and give them additional variable practice.” 

Poor use says: “The camera recorded a deviation, therefore the worker is at fault.” Good use says: “The video shows that the procedure is awkward, the labelling is unclear and the working position forces the person into a risky movement. We need to improve the conditions and the training.” 

In other words, technology should not only help assess the person. It should help improve the system. 

There is also an ethical question. If wearable devices, video analytics, voice assistants and digital traces of action are used, workers must understand what data is being collected, why, who has access to it, how it will be used, whether it may be used for disciplinary decisions and how privacy is protected. If people experience technology as a tool of punishment, they will hide mistakes, game the metrics and avoid honest feedback. If technology is embedded in a culture of learning, it can become a powerful tool for development. 

The principle should be simple: learning and safety first, control second. 

From course to skill-development trajectory 

Bernstein’s approach ultimately forces us to rethink the model of corporate training itself. The traditional model is: course, test, certificate. A stronger model is: task, practice, feedback, correction, variable scenarios, reassessment, transfer to real work and further development. 

In such a model, the certificate does not disappear, but it ceases to be the main proof of readiness. The development trajectory of the skill comes to the foreground. For example, a worker may have more than a note saying “completed equipment isolation training”. They may have a development map: knows the procedure; understands the purpose of isolation; performs the standard sequence; does not miss critical check points; acts correctly under changed configuration; recognises a mismatch between the drawing and the actual state; stops under uncertainty; can explain the risks to another worker; and performs reliably across different scenarios. 

That is much closer to real competence. 

If we accept Bernstein’s view, competence is no longer simply the sum of knowledge, procedures and training hours. It becomes the ability to construct reliable action under real conditions. It includes knowledge, task understanding, practical performance, cognitive diagnosis, use of feedback, correction, adaptation, communication, work under uncertainty, the ability to stop, the ability to learn from error and the ability to transfer skill to a new context. 

This is why classroom training remains necessary, but insufficient. The classroom can provide language, models, structure and procedural understanding. But practice forms action. Scenarios teach people to apply procedures under conditions. Feedback teaches them to correct their model. Variable cases teach them to think. Field coaching helps transfer the skill into real work. 

Nikolai Bernstein is rarely mentioned in corporate learning strategies, oil and gas training proposals, competency matrices or maintenance training manuals. Perhaps it is time to change that. 

His ideas remind us that the professional does not merely execute commands. The professional solves tasks. They use feedback. They adapt. They automate the right elements while retaining meaningful control. They learn through variability, correction and experience. 

For corporate training, this means the goal is not simply to transfer information. The goal is to form action. 

For HSE, it means that human error cannot be reduced to individual blame. The goal is to understand the conditions in which action is constructed. 

For operations and maintenance, it means that competence is not proven only by attendance, certificates or written tests. The goal is reliable performance in the real working environment. 

And for the oil and gas industry, where the consequences of error can be severe, this distinction is not theoretical. It is practical, operational and ethical. 

The forgotten science of skill has not been forgotten because it is obsolete. It has been forgotten because we have not yet fully translated it into the language of modern corporate training. 

It is time to do so.