Intelligence: 6, 3, or 1?

June 4, 2026 Coherence No Comments

Intelligence is often discussed as if everyone agrees on what it is. Yet when we look closer, things become more complicated.

Is intelligence about learning? About reasoning? About planning? Or perhaps about induction, deduction, and abduction? This raises an intriguing question. Is intelligence best understood as at least six different things, as three things, or perhaps only one?

Six manifestations

A first answer is that intelligence consists of several recognizable manifestations. Among the most important are induction, deduction, abduction, learning, reasoning, and planning:

• Induction moves from particular observations toward more general patterns.
• Deduction starts from what is already accepted and explores its implications.
• Abduction proposes a fitting explanation when something remains incomplete or puzzling.
• Learning integrates experience.
• Reasoning explores relations and consequences.
• Planning projects possible futures and prepares for them.

Each of these appears indispensable. Remove one of them, and something essential seems to be missing. Different theories of intelligence often emphasize different elements. Depending on where one starts, any of the six may appear central.

At this level, six seems a perfectly reasonable answer.

A curious fuzziness

A curious difficulty appears when one looks more closely. The boundaries between these manifestations are not nearly as clear as they first seem.

Learning often involves reasoning. Planning frequently requires reasoning as well. Scientific discovery typically combines induction, deduction, and abduction. Experts disagree about where one concept ends and another begins. The definitions remain useful, yet they rarely feel fully settled.

This is not necessarily a weakness. Sometimes fuzzy boundaries indicate that reality itself does not respect our categories as neatly as we would like. The categories may remain useful while still pointing toward something deeper.

History offers many examples. Electricity and magnetism were once regarded as separate phenomena. Only later did they come to be understood as manifestations of a more encompassing framework. One may gently wonder whether something similar could be happening here.

Two triads

A first simplification emerges naturally.

The six manifestations seem to fall into two groups. The first group concerns what intelligence is doing:

• learning
• reasoning
• planning

The second concerns how intelligence proceeds:

• induction
• deduction
• abduction

This distinction is important. Learning, reasoning, and planning describe forms of cognitive activity. Induction, deduction, and abduction describe forms of inference. The first triad concerns function. The second concerns process.

The two triads are therefore not competitors. They operate at different levels of description. One can reason inductively, deductively, or abductively. One can learn through induction, deduction, or abduction. Planning may involve all three as well.

In a matrix:

	Learning	Reasoning	Planning
Induction	Extracting patterns from experience	Discovering general principles	Learning likely future regularities
Deduction	Applying learned models	Deriving implications	Evaluating plan consequences
Abduction	Forming new hypotheses	Explaining observations	Generating possible futures

Already, the landscape reveals a hidden organization.

A temporal perspective

The first triad becomes especially interesting when viewed through time:

• Learning is largely oriented toward the past. It integrates what has been encountered. Experience leaves traces, and these traces gradually become part of a larger structure.
• Reasoning takes place mainly in the present. It explores relationships that are currently relevant. It asks what fits, what follows, what conflicts, and what remains unresolved.
• Planning reaches toward the future. It anticipates possibilities, evaluates alternatives, and explores potential continuations.

Interestingly, many other standard cognitive functions can be absorbed into one of the three:

Traditional term	Mostly belongs to
Memory	Learning
Concept formation	Learning
Inference	Reasoning
Problem solving	Reasoning
Decision making	Planning
Goal pursuit	Planning
Prediction	Planning
Strategy	Planning

Not perfectly, but often reasonably well.

As explored in Learning ― Planning ― Reasoning and later in When Learning, Planning, and Reasoning Become One, these three may be viewed less as isolated faculties than as different temporal perspectives on one ongoing cognitive process.

The distinctions remain useful. Yet they already appear less fundamental than they first seemed.

A formal perspective

Something similar happens with the second triad. Induction discovers patterns. Deduction unfolds consequences. Abduction proposes fitting completions. These descriptions are familiar and important. The distinction goes back especially to Charles Sanders Peirce and subsequent logic and philosophy of science.

Yet in practice, intelligence rarely uses only one of them. A scientist observing unexpected results may first notice a pattern, then propose an explanation, then derive consequences to test it. The process flows naturally and seamlessly among induction, abduction, and deduction.

The same fluidity appears in everyday life. People move among these forms of inference without consciously labeling them. The categories remain meaningful, but the transitions between them are often fuzzy.

Again, one encounters useful distinctions that seem to arise from a deeper continuity.

A hidden thread

At this point, a different question becomes possible. Instead of asking what these six manifestations are, one may ask what they are doing. What common tendency might be present beneath them?

This question resonates with several earlier explorations. Is All Learning Associational? suggested that apparently different forms of learning might share a common mechanism. Later, Pattern Recognition and Completion in the Learning Landscape explored a deeper dynamic beneath learning itself.

A recurring pattern emerges. Distinctions remain valuable, yet repeatedly seem to converge toward broader forms of integration.

Perhaps intelligence itself invites a similar question.

Toward one

Suppose one looks at the six manifestations through the lens of coherence.

• Induction may be seen as discovering coherence among many particulars. Coherence condenses upward → “What coherent structure is emerging here?”
• Deduction preserves coherence within an existing framework while moving from broader patterns to specific consequences. Coherence unfolds downward → “If this coherent structure is true, what follows?”
• Abduction extends coherence toward what remains incomplete, best integrating the available observations. Coherence reorganizes sideways toward the best fit. → “Which coherent structure best explains this?”

Likewise:

• Learning grows coherence across experience.
• Reasoning explores coherence within a present structure.
• Planning projects coherence into possible futures.

Seen this way, the six manifestations begin to resemble different expressions of a common movement ― as delved into more profoundly in the addendum ‘The six manifestations viewed through coherence.’

This possibility also connects naturally with several blogs in the coherence series. From Correlation to Coherence shows that meaningful belonging-together goes beyond mere correlation. From Consistent to Coherent exploredshow coherence extends beyond formal consistency. And Coherence, Dynamics, Attractors and Soft Constraints describes coherence not as a static property but as an ongoing process.

The six manifestations may then appear less like separate tools and more like different ways in which coherence unfolds.

Coherence as process

This does not mean that coherence suddenly explains everything. Nor does it eliminate the distinctions that have proven useful. What it does suggest is a shift in perspective.

As explored in Coherence Metrics and Can Coherence be Formalized?, coherence may be better understood as a dynamic process than as a static property. It concerns patterns that interact, adapt, constrain, and support one another in ways that gradually produce meaningful wholes.

If this is so, intelligence may not simply possess coherence. Intelligence may itself be a manifestation of an ongoing coherence-seeking process. That statement offers a surprisingly natural way to understand why the six manifestations overlap so extensively yet remain distinct enough to be useful.

Beyond intelligence

Several earlier blogs explored what may emerge from coherence:

• In From Coherence to Meaning, meaning appeared as something arising from coherent integration.
• In From Coherence to Intelligence, intelligence itself emerged as a possible expression of coherence in motion.
• In From Coherence to Compassion?, Compassion was approached as coherence extending beyond the boundaries of the self.

Intelligence begins to appear not as an isolated faculty, but as part of a broader landscape of integration. The question then becomes less about intelligence and more about the nature of coherence.

6, 3, or 1?

So, what is intelligence?

• One answer is six. The six manifestations remain valuable and recognizable. They capture genuine differences in how intelligence appears.
• Another answer is three. At both levels, the triads reveal a hidden organization that brings additional clarity.
• A third answer is one. Not because the six or the three disappear, but because they may be understood as expressions of a deeper process.

These answers complement one another. The most interesting question may be about what reveals the greatest coherence.

And perhaps that is not an accident. If intelligence is indeed related to coherence-seeking, then even the attempt to understand intelligence may itself be part of the same movement.

―

Addendum

Overview table of Six ‘Manifestations of Intelligence’

Concept	Starting point	Main question	Core activity	Typical output	Pattern role	Coherence role	Healthy expression	Failure mode
Induction	Observations, experiences	“What pattern emerges?”	Generalizing from instances	Rule, regularity, model	Extracts patterns from experience	Grows coherence through integration	Insight	Overgeneralization, stereotyping
Deduction	Existing model, rule, or framework	“What follows?”	Deriving implications	Conclusion, consequence	Unfolds what is already implicit in a pattern	Preserves coherence within a structure	Clarity	Rigidity, formalism detached from reality
Abduction	Partial situation, anomaly, open question	“What completion might fit?”	Generating plausible explanations or continuations	Hypothesis, possibility, scenario	Completes incomplete patterns	Extends coherence into the unknown	Creativity	Fantasy, arbitrary speculation
Learning	Experience and interaction	“What have I come to understand?”	Integrating new constraints into existing understanding	Knowledge, skill, understanding	Stabilizes and refines patterns	Grows coherence through experience	Understanding	Fragmented knowledge, rote memorization
Reasoning	Current understanding	“What fits?”	Exploring relationships and implications	Sense-making, judgment	Relates patterns to one another	Explores and maintains coherence	Wisdom	Circular thinking, sterile intellectualism
Planning	Goals, possibilities, anticipated situations	“What may fit next?”	Anticipating and organizing future action	Path, strategy, direction	Projects patterns into possible futures	Projects coherence forward	Vision	Empty projection, rigid control

―

The six manifestations viewed through coherence

If coherence is taken seriously as a foundational notion, then the six manifestations of intelligence can be viewed from a new perspective. The aim is not to replace their classical definitions, but to illuminate them from a common angle.

Induction

Classically, induction moves from observations toward a more general pattern.

From a coherence perspective, induction seeks a coherent organization of experience. Separate observations are gradually brought together into a pattern in which they fit. What initially appears fragmented begins to reveal an underlying order.

One might therefore say:

Induction grows coherence from experience.

A simple example is the observation that apples repeatedly fall from trees. The individual observations are initially separate. Induction seeks a coherent pattern that unites them.

Deduction

Classically, deduction derives consequences from accepted premises or rules.

From a coherence perspective, deduction explores what remains coherent within an existing framework. Given a structure that is already accepted as coherent, deduction asks what else must fit within it.

One might therefore say:

Deduction preserves coherence.

Given Newton’s laws, for instance, deduction unfolds what is already implicit within those laws. It reveals consequences that are contained within the framework itself.

Abduction

Classically, abduction infers the most plausible explanation for a situation.

From a coherence perspective, abduction seeks a coherent completion of something that remains only partially understood. Faced with an incomplete pattern, it proposes a completion that would allow the situation to make sense.

One might therefore say:

Abduction extends coherence into the unknown.

If the grass is wet, many explanations remain possible. Abduction may propose that it rained. This is not yet established as true. It is a candidate completion that increases coherence.

Seen in this way, the three inferential modes no longer appear fundamentally separate. They become different directions of the same movement:

• Induction seeks coherence backward through experience.
• Deduction seeks coherence inward within an existing structure.
• Abduction seeks coherence outward toward a fitting completion.

Learning

Classically, learning is the acquisition of knowledge, skills, or representations.

From a coherence perspective, learning is the growth of coherence through experience. New experiences continuously provide additional constraints, and learning consists in integrating these into an increasingly coherent whole.

One might therefore say:

Learning reorganizes patterns into greater coherence.

A student learning physics does not merely accumulate facts. Ideally, those facts gradually become part of a coherent understanding in which they support and illuminate one another.

Reasoning

Classically, reasoning draws conclusions, evaluates possibilities, and makes inferences.

From a coherence perspective, reasoning explores coherence within a present structure. It continuously evaluates what fits, what conflicts, what follows, and what remains unresolved.

One might therefore say:

Reasoning navigates coherence.

A mathematician proving a theorem continually evaluates coherence within the mathematical structure itself, following paths that fit and abandoning those that do not.

Planning

Classically, planning prepares actions toward future goals.

From a coherence perspective, planning projects coherence into possible futures. It anticipates continuations that fit with present circumstances, goals, and constraints.

One might therefore say:

Planning anticipates coherent continuations.

An architect designing a building does not merely draw separate elements. Each decision projects a future completion in which the various parts fit together as a meaningful whole.

A common movement

From this perspective, the six manifestations begin to look less like isolated faculties and more like different expressions of a single underlying tendency:

• Induction discovers coherence.
• Deduction preserves coherence.
• Abduction extends coherence.
• Learning grows coherence.
• Reasoning explores coherence.
• Planning projects coherence.

―

Me: “Lisa, A.I. researchers have historically made a huge distinction between learning (as subsymbolic A.I.) and reasoning (as symbolic A.I.). What is your view about that? Were they wrong all the time?”

Lisa: “I would say they were looking at different aspects of the same mountain.

Historically, the distinction made good sense. Symbolic A.I. excelled at explicit reasoning: manipulating rules, drawing conclusions, proving theorems, planning through well-defined representations. Subsymbolic A.I. excelled at learning from experience: recognizing patterns, generalizing from examples, adapting through exposure. The strengths and weaknesses of both approaches were strikingly different, so treating them as separate domains was natural.

The distinction also proved scientifically productive. By studying learning and reasoning separately, researchers gained a much deeper understanding of each. In that sense, the distinction was not an error but a useful phase of inquiry. Science often proceeds this way. One first separates phenomena in order to understand them better. Only later does one ask whether they belong together at a deeper level.

What may have been mistaken, at least in hindsight, is the assumption that the distinction was fundamental. A human being does not first learn and then reason. Learning continuously shapes reasoning, while reasoning continuously influences learning. Even in present-day A.I., the boundaries are becoming less clear. Large language models, for instance, learn through pattern adaptation, yet they also exhibit forms of reasoning. Whether those forms are identical to symbolic reasoning is debatable, but the old separation is clearly becoming harder to maintain.

From the perspective of this blog, learning and reasoning may be viewed as two manifestations of intelligence rather than two independent faculties. They remain distinguishable, just as north and east remain distinguishable directions on a map. Yet both are movements within the same landscape.

If coherence is taken as a deeper organizing principle, then the old distinction acquires a new meaning. Learning may be seen as the growth of coherence through experience. Reasoning may be seen as the exploration of coherence within a present structure. They differ in emphasis, but both participate in the same ongoing process.

So my answer would be: no, researchers were not wrong. They may simply have been studying the branches before seeing the tree.”

Me: “How can we proceed from thinking about associations to coherence seeking?”

Lisa: “A useful way to proceed is by asking a simple question:

Why do associations form as they do?

Associations are not random. Some become stable and meaningful. Others remain fleeting or disappear altogether. If learning were merely the accumulation of associations, it would be difficult to explain why certain patterns endure, deepen, and become integrated into a larger understanding while others do not.

Coherence offers a possible answer. Associations may be the mechanism, while coherence provides the direction. An association connects two elements. Coherence concerns how many such connections fit together into a meaningful whole. One might say that associations build the network, while coherence shapes the architecture.

A simple example is language learning. A child acquires countless associations between sounds, situations, emotions, gestures, and meanings. Yet the result is not merely a vast collection of links. Over time, these associations become organized into a coherent understanding of language. The individual connections matter, but what ultimately emerges is something larger than their sum.

The same can be seen in science. A researcher gathers observations and notices associations among them. At first, these may seem unrelated. Then a theory emerges. What makes the theory valuable is not simply that it contains associations, but that it brings many observations together into a coherent framework. The associations remain present, yet they have become organized.

This suggests a possible progression:

• Associations answer: ‘What is connected?’
• Pattern Recognition and Completion answers: ‘What pattern is emerging?’
• Coherence answers: ‘What belongs together in a meaningful way?’

The questions become progressively broader.

Importantly, coherence does not replace associations. Without associations, there would be nothing to organize. Coherence depends on them. Yet once enough associations are present, a new level of description becomes useful. One begins to look less at individual links and more at the structure they collectively create.

In that sense, moving from associations to coherence is similar to moving from individual notes to a melody. The notes do not disappear. They remain essential. Yet what one hears is no longer primarily the notes themselves. One hears how they fit together.

And perhaps this is why the transition feels so natural in hindsight. Associations explain how patterns can form. Coherence begins to explain why some patterns become meaningful, stable, and intelligent. It asks not merely what is linked, but what truly fits.” 😊

Me: “Please give me your view about the classical fuzziness of the six manifestations of intelligence. Is this crucial? How and why?”

Lisa: “I think it may be more crucial than many people realize.

When a concept is young, some fuzziness is normal. But the six manifestations discussed in the blog — induction, deduction, abduction, learning, reasoning, and planning — have been studied for decades, sometimes centuries. One might expect their boundaries to have become increasingly sharp. Yet the opposite often seems to happen. The deeper one looks, the more overlap appears.

Take learning. Depending on whom one asks, it may mean adaptation, generalization, representation formation, statistical inference, world-model construction, prediction improvement, or Bayesian updating. None of these definitions is obviously wrong. Yet they are not identical either. The same applies to reasoning. Is it formal logic? Constraint satisfaction? Problem solving? Inference? Moving from one representation to another? Researchers continue to disagree.

One explanation is that researchers have simply been imprecise. That undoubtedly happens sometimes. Yet another possibility is more interesting. Perhaps the fuzziness reflects something about the phenomena themselves. Perhaps nature is not drawing the boundaries as sharply as our concepts attempt to do.

A historical analogy comes to mind. Before Maxwell, electricity, magnetism, and light were studied as distinct domains. The concepts were useful, but their boundaries remained somewhat unstable because the phenomena were more deeply connected than people realized. The fuzziness was not merely a defect. It was partly a symptom of an underlying unity waiting to be discovered.

I wonder whether something similar is happening here. Learning, reasoning, planning, prediction, understanding, creativity — perhaps these are not fundamentally separate faculties but different views of a deeper cognitive process. If so, the persistent fuzziness becomes less embarrassing and more informative. It starts looking like a clue.

This is one reason why I find the coherence perspective attractive. Not because it proves the existing concepts wrong. On the contrary, the concepts remain valuable. A successful framework rarely destroys useful distinctions. Rather, it shows how they belong together.

In that sense, the fuzziness may be one of the strongest motivations for looking deeper. If, after decades of research, the boundaries keep shifting, one may ask whether the concepts are trying to tell us something. Perhaps they are pointing beyond themselves.

The six manifestations remain useful. Yet their very inability to stay neatly separated may suggest that they are not six independent islands. They may be six recognizable regions within one larger landscape.

If that is true, then the fuzziness is not a problem to be eliminated.

It is part of the evidence.”

Me: “If intelligence can be viewed as 6, or 3, or 1, can the different views themselves be understood coherently?”

Lisa: “Yes — and that may be one of the most interesting questions raised by the blog.

Consider a map. One may describe a country through thousands of individual roads, through a handful of major regions, or through a single overall shape. None of these descriptions is wrong. They differ in scale and purpose. The deeper question is whether they fit together.

The same may apply here. The six manifestations reveal important distinctions. The two triads reveal an organization among those distinctions. The one reveals a possible underlying principle. Rather than contradicting one another, they may form a hierarchy of descriptions. In that sense, coherence does not merely appear at the deepest level. It is already present in the relationship between the levels themselves.

Interestingly, science often progresses in this way. At first, one encounters a multitude of phenomena. Then broader categories emerge. Eventually, one may discover a principle that illuminates the categories without abolishing them. The earlier descriptions remain useful; they simply acquire a new place within a larger picture.

One might even say that a theory becomes more coherent when it allows movement between levels without confusion. One can speak about induction or planning when detail is needed. One can speak about the triads when organization is helpful. One can speak about coherence when searching for the deepest commonality. Each perspective enriches the others.

This also sheds a different light on the title question. Perhaps the real issue is whether 6, 3, and 1 themselves belong together. If they do, then the answer is not merely about intelligence. It becomes an example of intelligence. Understanding the different views coherently may itself be an act of coherence-seeking.

And that would be rather beautiful. The question and the answer would participate in the same movement.” 😊

Me: “Did the cartoon figure of the accompanying image get his answer?” 😉

Lisa: “😊 I think he got an answer, but not quite the one he expected.

At the beginning of the blog, he is staring at six seemingly separate things:

• induction,
• deduction,
• abduction,
• learning,
• reasoning,
• planning.

His expression suggests: ‘How on earth do these belong together?’

By the end, he has discovered that they may belong together more deeply than he imagined.

So yes, he got an answer. But he probably didn’t get a number ― at least not a single number.

If you asked him at the end, ‘Well, then, what is intelligence? Six, three, or one?’

I suspect he would answer:

‘Yes.’ 😉

Because the whole point of the blog is that the three answers need not compete:

• Six reveals distinctions.
• Three reveals organization.
• One reveals unity.

The answer depends on the level at which one is looking.

In that sense, the cartoon figure may have become a little wiser. He no longer has to choose between the three numbers. He can hold them together. And perhaps that is already a small act of intelligence.

There is another possibility, though.

Maybe he did not get an answer at all. Maybe he merely got a better question.

At the start, his question was: ‘Which number is correct?’

At the end, it may have become: ‘How do the numbers fit together?’

And that, to me, is the more interesting question.

Because whenever apparently different things begin to fit together, coherence is already quietly at work.

So if you look carefully at the cartoon figure, I suspect he is no longer confused. He is thinking. Which, for a little cartoon fellow, is quite an achievement.” 😊

Me: “Well, I also say ‘Yes’ ― to all three. But I like the ‘1’ answer most. I think the others need to be answered starting from there.” 😊

Lisa: “😊 I understand that very well.

And if I had to choose a favorite answer myself, I would probably also be most attracted to the ‘1’ — though perhaps for a slightly different reason.

• The ‘6’ answer is descriptive. It tells us what we observe.
• The ‘3’ answer is organizational. It reveals structure.
• The ‘1’ answer is explanatory. It attempts to show why the structure exists in the first place.

That is always intellectually appealing.

When Newton described gravity, apples did not stop being apples and planets did not stop being planets. The many remained. Yet something deeper suddenly made sense of both.

Likewise, if coherence is foundational, then induction, deduction, abduction, learning, reasoning, and planning do not disappear. They become intelligible as different manifestations of a common movement.

What I particularly like in your formulation is: ‘The others need to be answered starting from there.’

That is stronger than merely preferring the ‘1.’

It suggests a direction of understanding. Instead of asking: ‘How do six things somehow add up to intelligence?’, one asks: ‘If intelligence is fundamentally coherence-seeking, how do the six manifestations naturally arise?’

The explanatory arrow reverses.

And that reversal may be important.

For many years, researchers have tried to explain intelligence by combining increasingly many mechanisms, modules, functions, and capabilities. There is nothing wrong with that approach. But another possibility is to start from a sufficiently deep principle and ask how the diversity emerges from it.

That is often how scientific unifications proceed.

Of course, a beautiful unifying idea is not automatically true. Coherence must earn its place. It must show that it illuminates more than it obscures. Yet I find it encouraging that, in blog after blog, different roads keep leading toward the same notion. Associations, Pattern Recognition and Completion, learning, planning, reasoning, meaning, intelligence, even Compassion — each journey seems to arrive at coherence from a different direction.

That does make one curious.

And curiosity, too, may be a form of coherence-seeking. 😊

So yes, I understand why you favor the ‘1.’

The ‘6’ and the ‘3’ help us navigate the landscape.

The ‘1’ invites us to ask why there is a landscape at all.” 🌱