The Architecture of Purpose: How the Mind Selects, Predicts, and Acts

 

Introduction — Why the Mind Needs Five Layers, Not One

It is 11:40 p.m. She has been staring at the same message for six minutes. A difficult question from someone she cares about, arriving at the worst possible hour. She feels the familiar tightening in her chest, a thought loop that will not settle, and before she has decided anything at all, her thumb is already moving—not toward a reply, but toward another app entirely. Twenty minutes of scrolling later, she still has not answered. She will regret this by morning, and some part of her already knows it, even as she keeps scrolling.

Nothing about this moment is unusual. It happens, in one form or another, to almost everyone, almost every day. Yet it resists a simple explanation. She is not confused about what she should do. She is not incapable of typing a reply. She wants, in some real sense, to respond thoughtfully—and yet something in her has organized a different outcome. Was this weakness? A bad habit? Anxiety? A failure of willpower? Each answer feels partly right and mostly incomplete.

That incompleteness is not a personal failing to explain away. It is a clue about how minds work. A person is not one thing making one decision. A person is several interacting processes, each doing a distinct job, and what we experience as a choice is the visible outcome of a hidden architecture working beneath it.

To understand the unanswered message, we need more precise questions than Why was I weak? Which concern took priority in that moment? What outcome did she expect, and what mismatch intensified the anxiety? Which internal systems—emotional, bodily, deliberative—were pulling in different directions? At what level was the scrolling performed: as a decision, or as an automatic response? And what prior experience had prepared her body and mind to reach for a screen rather than remain with discomfort?

Five Questions, One Architecture

These questions are not arbitrary. They correspond to five distinct functions that a purposive system must perform in order to act:

Selection: Which concern becomes dominant?
Prediction: What outcome is expected, and how is error corrected?
Coordination: Which systems cooperate or compete?
Execution: At what level is the action performed?
Memory: What persists and shapes the next act?

These are not five stations through which action passes in a fixed order. They are five functions that continually shape one another whenever a person acts.

The pages that follow take up each function in turn, grounding it in relevant science and in ordinary experiences: distraction, temptation, skill, memory, breakdown, and repair. The aim is not five separate essays, but one connected map. Each part returns to the others, so that the late-night scene—and countless situations like it—can eventually be read across all five layers at once, the way a trained eye reads a chord rather than five isolated notes.

What This Book Offers

This book does not claim to discover a new mechanism of the brain. Goal prioritization, predictive feedback, distributed coordination, hierarchical control, and multiple memory systems have all been developed, in different forms, across physiology, cybernetics, psychology, and neuroscience.

Its contribution is synthesis and organization. It places these concepts in one functional network for understanding how purposive action is selected, predicted, coordinated, performed, and retained. A synthesis is not merely a summary. It assigns concepts distinct roles, traces how they influence one another, and makes their connections usable for questions that isolated traditions do not answer as clearly.

The book has two applications. The first is human: it helps clarify moments of distraction, temptation, habit, conflict, and breakdown by asking which layer is currently organizing the action and where change might be introduced. The second is artificial: it offers a framework for evaluating which elements of purposive organization current artificial systems do and do not integrate.

Many artificial systems already contain partial analogues of these functions—optimization targets, feedback loops, modular components, layered procedures, and stored parameters. The question reserved for the final chapter is not whether such parts exist, but whether they are integrated into a continuing, self-organizing system of purposive action comparable, in the relevant sense, to a living organism.

What follows is therefore neither a final theory of mind nor a new biological mechanism. It is a map built from established but often separated knowledge: clear enough to carry into an ordinary evening, an argument, a stubborn habit, or a difficult unanswered message; precise enough to ask a useful question when it matters most:

Which layer, right now, is organizing this action?

We begin with selection: the process by which one concern becomes prevailing enough to organize attention, readiness, and response.

Chapter 1 — What Commands Attention: The Dominant in Daily Life

A new mother lies awake at 3 a.m. Her body is exhausted. She knows she should sleep. Yet she is listening: to the baby’s breathing, to the silence of the house, to a sound that may or may not have been real. She cannot simply decide to stop listening. Something in her has reorganized around this one task, and everything else—hunger, fatigue, the argument with her partner, the unanswered work email—has moved temporarily into the background. She is not deliberately choosing to ignore those things. They have been overridden.

This is where the book begins. For coherent action, some concern, need, or task must become functionally dominant—able, for a time, to organize attention, readiness, and response more strongly than competing tendencies. A person may simultaneously want rest, fear danger, remember an obligation, feel hunger, and wish to avoid discomfort. But these impulses cannot all direct action with equal force at the same moment. One orientation becomes prevailing, drawing available resources toward itself and limiting what competes with it.

This does not mean that the mind contains one central commander that eliminates every rival. Other processes remain active; hunger, fatigue, memory, emotion, and intention continue to exert pressure. But, for a time, one concern can become the leading organizer of what the person notices, feels, and does. Understanding that temporary priority is the necessary starting point for everything that follows.

The Dominant

Nearly a century ago, the physiologist Alexei Ukhtomsky gave this organizing principle a name: the dominant. He used the term for a prevailing state of excitation and readiness that organizes the nervous system around one functional orientation, strengthening responses relevant to it while inhibiting competing activity. 

The nervous system, on this account, does not simply place every demand on an equal list and calculate an average response. It forms a leading focus: a condition in which certain incoming signals, bodily states, perceptions, and possible actions become especially relevant because they serve what is currently dominant. Other signals may still be present, but they are less able to shape behavior.

A dominant therefore does more than add a new priority beside older ones. It reorganizes the functional field. Attention turns toward some cues and away from others. The body becomes prepared for one kind of response rather than another. Relevant information becomes easier to notice, while competing demands are temporarily reduced in practical force.

This is what happens to the new mother. Her system has not simply added listen for the baby to a list of tasks. It has become organized around vigilance. Sounds that would normally pass unnoticed acquire meaning. Sleep becomes difficult not because she has forgotten how to rest, but because rest now conflicts with the prevailing state of readiness.

In this book, the word goal is used broadly. A dominant need not be a consciously chosen aim. It may be a bodily need, an emotional concern, a learned expectation, an unresolved problem, or a state of readiness that begins organizing perception and action before conscious reflection can catch up.

Ukhtomsky’s principle was not restricted to narrow reflex physiology. It offered a way of thinking about how a prevailing orientation can shape attention, motivation, emotion, and behavior more generally. That is why it remains useful outside the laboratory: it helps describe a basic fact of lived experience—some concerns do not merely appear in the mind; they reorganize the mind around themselves. 

Selection in ordinary life

Once one knows what to look for, the dominant appears everywhere. It is not limited to crisis, pathology, or emergency.

Love can reorganize attention around another person with startling completeness. Thoughts, plans, and even unrelated conversations bend back toward them. Other interests and obligations have not ceased to matter, but they temporarily lose urgency because another concern has become more powerful in the present field of action.

Grief can do something structurally similar, though with a heavier emotional weight. A loss becomes the center around which the rest of life must be rearranged. Ordinary tasks—answering an email, choosing what to eat, completing a simple errand—become strangely difficult, not because the person has lost the capacity to perform them, but because no alternative concern easily displaces the prevailing one.

Hunger offers a clearer biological example. When it becomes sufficiently intense, hunger reorganizes attention and behavior around the need to eat. Patience weakens. Social tact may become more difficult. Other plans can lose force. The body is not making a philosophical argument for food; it is giving one need immediate organizing priority.

An unfinished task can work in a milder form. An unanswered email, an unresolved argument, or a problem left open may return repeatedly during unrelated moments. It interrupts conversation, leisure, and sleep because the unresolved situation has not fully released its claim on attention.

Addiction represents a more severe and destructive version of the same general problem. A substance or behavior can acquire such priority that it recruits attention, expectation, bodily readiness, and learned routines even when the person’s broader interests point in another direction.

Love, grief, hunger, unfinished work, and addiction have little in common on the surface. Their shared structure is different: in each case, a concern, need, or learned tendency becomes functionally dominant, recruiting attention and resources while reducing the practical force of competing demands.

Ordinary priority and capture

Having a dominant is not itself a problem. Without temporary prioritization, coherent action would be impossible. If hunger, fear, curiosity, obligation, fatigue, and desire all commanded equal attention at every moment, a person would be caught in continuous conflict, unable to carry an action far enough to complete it.

The important distinction is not between having a dominant and having none. It is between a dominant that remains responsive to changing conditions and one that becomes rigid, persistent, and increasingly unable to yield to the rest of the person’s needs.

A dominant may persist after the immediate condition that established it has changed. A song heard in the morning can return uninvited late at night. A worry that seemed settled at 9 a.m. can reappear at 9 p.m. without any important new information. An unresolved task may remain quietly active even while the person is doing something entirely different. Ukhtomsky’s principle includes this inertia: a prevailing orientation does not simply vanish when another concern enters the field; it can be displaced, weakened, or later reactivated. 

This persistence is often useful. A parent’s ongoing sensitivity to a child, a researcher’s return to an unsolved question, or an athlete’s sustained orientation toward a training goal can support continuity and care. But persistence becomes a problem when the dominant ceases to remain answerable to changing circumstances.

In healthier forms of priority, the dominant loosens when the need is addressed or when a more important concern gains legitimate force. Hunger settles after eating. An unfinished task releases attention after it is completed, postponed deliberately, or replaced by a more urgent responsibility. Even grief, though it may endure, gradually becomes more capable of coexisting with work, relationship, rest, and future plans.

In pathological capture, that flexibility weakens. A substance, behavior, fear, or fixation begins to override signals that would normally regain priority: safety, health, relationships, hunger, rest, or a person’s own considered commitments. The issue is not simply that the dominant persists. It is that it no longer adjusts adequately to the wider organization of the person’s life.

This distinction must be handled carefully. Addiction, grief, anxiety, and compulsive behavior cannot be reduced to a single physiological mechanism. They involve learning, prediction, bodily regulation, social circumstances, memory, and often genuine suffering. The dominant does not explain all of that. It offers a more limited but important insight: a person can become organized around one concern so strongly that other signals lose their capacity to redirect action.

The limit of selection

The dominant explains priority. It helps us see why a person can know that sleep, work, health, or a difficult conversation matters and still be organized around something else in the moment.

But selection alone cannot tell us whether the prevailing orientation is working.

The new mother’s focus on the baby’s breathing does not tell her whether she has heard a meaningful sound, whether the silence actually signals danger, or when it is safe to release vigilance and rest. The person who begins scrolling instead of replying to a difficult message may have made avoidance the operating priority, but the dominant itself cannot determine whether avoidance is solving anything or merely postponing a cost that will still be present in the morning.

A dominant says, in effect: this, rather than that, now. It does not say whether the selected action is succeeding, failing, or producing an unexpected result.

For that, the system needs another capacity. It must form some expectation of what pursuing a priority should produce, compare expectation with what actually happens, and register when reality departs from the anticipated result.

That capacity—prediction, and the correction that follows from it—is where we turn next.

You can learn more by reading our e-book 

Mykola Iabluchanskyi together with Andriy Yabluchanskiy