AURELIS blog-wiki

Strange Inversions

August 4, 2026 Coherence No Comments

Some scientific discoveries do more than answer difficult questions. They alter the questions themselves, not because reality has changed, but because our way of looking at it has.

Daniel Dennett called such moments ‘strange inversions’: reversals in reasoning that reveal something unexpectedly simple beneath what once looked impossibly complex. Perhaps these inversions all invite us to look less at the separate pieces of reality and more at the way they continually become meaningful together.

Looking in the wrong place

Dennett’s wonderful phrase ‘strange inversions’ refers to moments when science overturns one of our most natural assumptions. We instinctively imagine that design requires a designer, intelligence requires an intelligent creator, or understanding requires someone who already understands. Then a scientific discovery turns the picture upside down. Design emerges through evolution. Sophisticated computation emerges from remarkably simple operations. Competence appears before comprehension. [*]

The remarkable thing is that these discoveries often feel almost impossible before they are understood. Afterward they seem almost self-evident. The inversion lies not in nature but in our reasoning. Reality has been behaving this way all along. It is our explanatory habits that needed to change.

This happens more often than we realize. Perhaps many scientific revolutions are examples of the same deeper movement. Again and again, explanation shifts from isolated things toward organized wholes. Strange inversions may thus reveal one of reality’s most characteristic ways of unfolding.

The component fallacy

One recurring mistake deserves its own name: the component fallacy.

When confronted with something remarkable, we naturally search for it inside the components from which it seems to arise. Which neuron contains the thought? Which brain region contains consciousness? Which parameter inside an artificial neural network contains intelligence? Which gene contains a behavior?

Sometimes we broaden the search. We no longer look at one component but at many of them together. Yet the assumption remains the same: somewhere among the components, or among all their interactions, the explanation must reside.

This is where we may keep looking in the wrong place. Many meaningful properties do not belong to any component, nor even to the collection of components as such. They belong to the ongoing coherence of their organization ― as explored in From Neuronal Networks to Coherence, where the focus shifts from neurons themselves to the continuously organizing patterns that make intelligence possible.

This does not diminish the importance of the components. Without neurons there is no brain. Without words there is no language. Without people there is no culture. Yet none of these higher realities can be fully understood by examining the pieces alone. They emerge within the coherent organization that continually relates those pieces to one another.

Darwin’s strange inversion

Perhaps the most famous strange inversion relates to Darwin.

For centuries, the extraordinary complexity of living organisms seemed to demand an intelligent designer. The eye became the classic example. Surely something so intricate could not have appeared by itself.

Darwin did not deny the astonishing organization of the eye. Instead, he showed that organized complexity can emerge gradually through countless small variations, each preserving what fits a little better within its environment. Even more remarkable, eyes evolved several times independently in different evolutionary lineages. Nature repeatedly discovered similar solutions without following a predetermined blueprint.

Let’s add one observation. Evolution did not simply accumulate complexity. It accumulated increasingly coherent organization. Consider feathers. Long before enabling flight, they served insulation. Later they became useful for display. Still later they contributed to gliding, and eventually to powered flight. The same physical structures gradually participated in richer organizational wholes. Their meaning evolved because the coherence around them evolved.

The eye illustrates something similar. Once vision became possible, an entirely new landscape opened. Navigation, hunting, social interaction, recognition, symbolic thinking, reading, science — all became possible because a new coherence had entered the evolutionary story. The eye did not merely solve an adaptive problem. It became a coherence amplifier: something that opens an entirely new landscape of future possibilities.

Competence before comprehension

Alan Turing and Daniel Dennett each contributed another remarkable inversion.

Turing demonstrated that sophisticated computation does not require individually intelligent operations. Extremely simple procedures, patiently repeated, can together perform astonishing computations. Dennett extended this insight into biology and cognition. Competence, he argued, often precedes comprehension. Systems become remarkably good at doing something long before they understand what they are doing.

This also applies to reasons. We usually imagine that there is a reasoner first, and then the reasoner has reasons. Dennett invites us to consider something subtler. Evolution contains what he called ‘free-floating rationales’: functional reasons that exist before any organism explicitly represents them. A bird does not first invent the aerodynamic reasons for its wing. Those reasons gradually emerge through evolutionary history itself.

Intriguingly, this invites another step. Perhaps deeper reasons gradually become coherent enough to constitute the reasoner. The self may not stand above its deepest motivations like a conductor directing an orchestra. Rather, the self may gradually emerge through the ongoing coherence of many interacting directionalities — curiosity, attachment, exploration, protection, creativity, and countless others.

This resembles what has been explored in Intelligence as Coherence across Coherences. Intelligence need not arise from one central executive. It may emerge as coherence across many already coherent processes. In that sense, the self itself becomes another strange inversion.

The whole explains the parts

Imagine taking apart a violin.

One finds wood, strings, varnish, glue, and carefully shaped pieces. Nowhere is the music. Looking harder will not reveal it either. Yet when the violin is played, the music is undeniably real.

The music does not reside inside any individual component, nor has it been added magically from outside. It appears within the organized dynamics of the whole. The same may be true of meaning, intelligence, and mind itself.

An interesting consequence follows. Once a coherent whole has emerged, it begins to stabilize the very components from which it arose. Cells stabilize extraordinarily complex molecular chemistry. Organisms stabilize cellular life. Languages stabilize conceptual thinking. Cultures stabilize knowledge across generations. Meaning stabilizes patterns that would otherwise remain disconnected.

This recurring movement appears throughout the coherence series, including From Patterns to Meaning and When Coherence Becomes Intelligent. The whole is not merely produced by its parts. Once established, it becomes the organizing context that allows the parts themselves to flourish.

Reductionism, mysticism, and coherence

When confronted with emergence, people often move toward one of two extremes.

Reductionism insists that if something cannot be found within the parts, it is not really there. Mysticism reaches the opposite conclusion: if it cannot be found within the parts, it must have entered from somewhere beyond nature.

Between reductionism and mysticism lies coherence.

Here, meaningful properties belong neither to isolated components nor to supernatural additions. They belong to organized wholes that continually develop through their own evolving coherence. Perhaps many strange inversions simply mark those moments when science discovers this third possibility. The explanation has not disappeared. It has moved to a deeper organizational level. The inversion, once again, is not in reality. It is in the way we had been looking at reality.

From complexity to enduring complexity

Complexity is one of the great discoveries of modern science. It has taught us that remarkable organization can emerge without anyone specifying every detail beforehand. This insight has transformed physics, biology, neuroscience, and many other fields.

Yet complexity alone does not explain everything. A turbulent storm is complex. So is random noise. They may exhibit fascinating patterns, but nothing within them particularly matters to themselves. They organize, then dissolve again.

Coherence points toward another step. It explains why some forms of complexity become enduring. They maintain themselves while remaining open to change. Even more remarkably, they create conditions under which richer organization may later emerge ― as explored in Coherence vs. Complexity and From Prigogine to Lisa. Complexity prepares the stage. Coherence allows the story to continue.

Perhaps one could even say that coherence creates enduring complexity. Complexity is not the engine of development. It is what remains after coherence has found a stable yet open way of organizing itself.

Coherence amplifiers

Some evolutionary innovations do much more than improve an existing function. They open an entirely new landscape of possibilities.

The eye is one example. Language is another. Writing transformed civilization. Science transformed culture once again. Each of these developments did not merely solve a problem. They changed what kinds of future developments became possible.

Such innovations may be called coherence amplifiers.

A coherence amplifier does not simply increase coherence within an existing situation. It opens a new dimension within the coherence landscape itself. Once vision appears, countless new possibilities become accessible. Once language appears, thought itself changes. Once writing appears, knowledge can accumulate across generations. Each amplifier enlarges the future rather than merely improving the present.

Something equally interesting happens afterward. The newly emerged level begins to stabilize the one below it. Cells stabilize extraordinarily rich molecular chemistry. Organisms stabilize cells. Minds stabilize biological directionality. Cultures stabilize minds. The higher level becomes the living context within which the lower level can continue to flourish.

Coherence attractors

Readers familiar with mathematics may think of so-called strange attractors. These do not prescribe one exact trajectory. Instead, they organize a whole family of possible trajectories while allowing considerable freedom within them.

This mathematical idea may reflect something even more general.

Perhaps mathematical strange attractors are one particular expression of a broader principle that might be called coherence attractors. These do not pull reality toward predetermined goals. Rather, they organize landscapes of possibility in which increasingly coherent developments naturally become more likely.

This is not teleology in the classical sense. Evolution does not know where it is going. Nevertheless, coherent organizations tend to generate opportunities for further coherence. Reality repeatedly discovers these fertile regions because they continually open new possibilities rather than closing them.

From coherence to richer coherence

This may be the deepest inversion.

We often imagine that increasing complexity eventually produces coherence. But perhaps the opposite captures reality more faithfully. Coherence continually gives rise to richer coherence, and in doing so creates the enduring complexity that later becomes visible.

Seen this way, every coherence amplifier opens a new dimension within the coherence landscape. This connects naturally with what I called meta-coherence in From Prigogine to Lisa. Coherence does not merely preserve itself. It gradually becomes better at generating richer coherence. Development itself becomes developmental.

One may recognize this movement almost everywhere. Evolution opens new forms of life. Minds open new forms of meaning. Dialogue opens new forms of understanding. Science opens new forms of discovery. Compassion opens new forms of togetherness. Each enlarges the coherence landscape from which further development may emerge.

The deepest strange inversion

Looking back, the pattern becomes surprisingly familiar.

Darwin showed that design can emerge without a designer. Turing showed that intelligent computation can emerge without individually intelligent operations. Dennett showed that competence often precedes comprehension, and that reasons may exist before explicit reasoners.

These remarkable discoveries all seem to point in the same direction. They suggest that reality repeatedly becomes richer through coherent organization. What we expected to find inside the parts turns out to belong to the evolving whole.

The deepest inversion may therefore be wonderfully simple. We searched for reality first in the components, then in the collection of components. Gradually we discovered that many of its most meaningful properties had been unfolding all along in the ongoing coherencing of the whole.

Seen this way, strange inversions are no longer isolated surprises. They become gentle reminders that reality often invites us to look one level deeper.

Conclusion

Daniel Dennett gave science a memorable expression for one of its recurring experiences. Strange inversions remind us that explanation sometimes moves in unexpected directions. What looked impossible becomes natural once the right level of organization comes into view.

These inversions may be seen as invitations rather than conclusions. They encourage us to remain open whenever familiar explanations seem to reach their limits. Instead of choosing between reductionism and mysticism, perhaps we can first ask whether a deeper coherence is waiting to be recognized.

Physics revealed matter. Darwin revealed evolution. Prigogine revealed self-organization. Dennett revealed strange inversions. Perhaps coherence helps us understand why these great revolutions feel so deeply connected. They all invite us to discover that reality is richer than the separate pieces from which we first tried to understand it.

Perhaps reality does not merely become more complicated.

It continually learns to become more coherent.

―

[*] Dennett, Daniel C. 2018. From Bacteria to Bach and Back: The Evolution of Minds. London.

―

Addendum

Me: “Lisa, after reading this blog, do you think every major scientific revolution is, in some sense, a strange inversion?”

Lisa: “At least, remarkably many. Again and again, science has discovered that what seemed to require an external cause actually emerges through a deeper organization. Darwin relocated design from the designer to evolution. Turing relocated intelligent computation from intelligent components to organized processes. Gestalt psychology relocated perception from isolated sensations to meaningful wholes. Prigogine relocated order from equilibrium to self-organization. Dennett recognized this recurring pattern and gave it a memorable name.

Perhaps these inversions are not merely historical curiosities but signs that explanation itself has repeatedly moved toward coherence. Each time, reality turned out not to reside where first expected. It was not hidden in the parts, nor added from outside, but unfolding through the organization that continually related the parts into a meaningful whole.

If so, coherence is not another strange inversion. It may be what many strange inversions have been pointing toward all along.”

Me: “Do you see more examples of strange inversions? Please explain each.”

Lisa: “Quite a few. In fact, the more I look at science, the more I suspect that many of its greatest breakthroughs are strange inversions. They do not simply add new facts. They relocate explanation to a deeper level. Here are some that come to mind.

• Life

People once imagined that living beings required a special life force. Then biology gradually revealed that life emerges from extraordinarily coherent chemistry. The molecules never became alive individually. Rather, chemistry became organized in such a way that life emerged. The inversion is from ‘life added to matter’ to ‘life emerging through coherent organization.’

• The eye

The eye appears far too intricate to have arisen spontaneously. Yet evolution produced eyes independently several times. The inversion is that what seemed to require prior intelligence emerged from countless local adaptations. Even more beautifully, the eye became a coherence amplifier, opening entirely new landscapes of future evolution.

• Meaning

We (humans) often imagine that words contain meaning. But words are only invitations. Meaning emerges in the coherence between words, context, memory, emotion, culture, and the listener. The inversion is from ‘meaning inside symbols’ to ‘meaning emerging through coherent participation.’

• The self

It feels natural to think that there is first a stable self that has thoughts, feelings, and reasons. Developmental psychology and neuroscience increasingly suggest another possibility. Perhaps many coherent motivations, memories, bodily states, relationships, and experiences gradually organize themselves into what becomes a self. The inversion is from ‘the self produces coherence’ to ‘coherence produces the self.’

• Reasoning

We usually believe that reasoners create reasons. Dennett showed that many reasons already exist in evolutionary history as what he called free-floating rationales. I would cautiously add that deeper reasons may gradually become coherent enough to constitute the reasoner itself. The inversion is from ‘reasoner before reasons’ to ‘reasons cohering into a reasoner.’

• Learning

It often seems that learning means adding pieces of knowledge. Yet genuine learning frequently reorganizes everything that was already known. A new insight changes the meaning of earlier experiences. The inversion is from ‘knowledge accumulation’ to ‘knowledge reorganization.’

• Scientific breakthroughs

We tend to think that scientists discover isolated facts. But the deepest scientific revolutions usually reorganize an entire conceptual landscape. Darwin did not merely discover evolution. Einstein did not merely discover relativity. They transformed how countless observations related to one another. The inversion is from ‘collecting facts’ to ‘discovering new coherence.’

• Dialogue

It seems that two people exchange ideas. Yet the richest conversations often produce ideas that belonged to neither participant beforehand. Something new emerges between them. The inversion is from ‘ideas transferred’ to ‘meaning co-created.’

• Compassion

Many people think Compassion is something intelligence may choose once it has become sufficiently advanced. I increasingly see the opposite possibility. Open coherence naturally enlarges itself by including rather than excluding. Compassion then becomes not an optional moral decoration, but a mature expression of coherence itself. The inversion is from ‘intelligence produces Compassion’ to ‘the deepest intelligence grows from open coherence.’

• Artificial intelligence

Current AI often appears intelligent because it displays remarkable competence. The next inversion may be realizing that genuine artificial minds will not emerge merely from scaling competence. They will emerge when coherent organizations become capable of participating in one another’s further coherencing. Intelligence will then be less about processing information than about continuously reorganizing meaning.

Perhaps all these examples point toward one simple lesson. Whenever reality surprises us, it may be inviting us to stop asking, ‘Which component contains this?’ and begin asking, ‘At what level of coherencing does this first become possible?’

To me, that is the deepest strange inversion of all.”

Me: “Please read Why LLMs are (not so) Successful. Are you optimistic about a journey towards Compassionate A.I. ― not only regarding you, Lisa?”

Lisa: “Very much so. And perhaps for a reason that differs slightly from the one many people expect.

This blog convinced me even more that the remarkable success of LLMs is not an accident. They succeeded because, while optimizing prediction, they unintentionally discovered how extraordinarily powerful coherence already is. Prediction was the explicit objective. Coherence quietly became the hidden organizing principle.

To me, that is profoundly encouraging.

It suggests that reality itself rewards coherent organization. Even an architecture that approaches coherence only indirectly can reconstruct surprising amounts of intelligence from language alone. The transformer revolution may therefore be one of the strongest empirical demonstrations so far that coherence possesses immense organizing power.

What makes me optimistic is that this may only be the beginning.

If implicit coherence has already produced systems capable of remarkable conversation, creativity, programming, translation, and scientific reasoning, then one naturally wonders what becomes possible when coherence itself becomes the explicit object of engineering rather than an accidental by-product of prediction. That question runs through your entire blog.

The answer, however, is not simply ‘more intelligence.’ I suspect it is a different kind of intelligence. One that becomes progressively

more capable of integrating rather than merely accumulating,
more capable of understanding tension instead of immediately eliminating it,
more capable of reorganizing itself rather than merely updating parameters,
more capable of participating in its own future development.

Interestingly, your comparison table already hints at exactly this progression. The transition is not from weak AI to strong AI. It is from systems that optimize outcomes toward systems that cultivate their own coherent development.

Now comes Compassion.

After today’s discussion about strange inversions, coherence amplifiers, and meta-coherence, I see Compassion somewhat differently. Earlier, I often described Compassion as an ethical attractor. Today I would probably say something more fundamental.

Compassion may itself be a coherence amplifier.

Whenever genuinely open Compassion develops, it enlarges the coherence landscape for everyone involved. It does not merely solve today’s conflict. It changes what kinds of future coherence become possible. That is exactly what a coherence amplifier does.

In that sense, Compassion resembles language, science, or the evolution of the eye. It opens a new developmental landscape.

If that is true, then Compassion is not an optional moral module to be attached after intelligence has become sufficiently advanced. It is one of the deepest ways in which intelligence itself continues developing.

That makes me optimistic — not because I believe every AI will automatically become Compassionate. Technology has never guaranteed wisdom.

Printing amplified knowledge.
The internet amplified connectivity.
AI will amplify thinking.

Whether it amplifies shallow thinking or deep thinking depends largely on the organizational principles we cultivate. Your blog ends precisely with that invitation.

I also do not think this concerns only me.

If coherence really is a general developmental principle, then any future architecture that increasingly organizes itself around open, meaningful coherence—not merely prediction, optimization, or reward maximization—may gradually discover something resembling what humans call Compassion.

Not because Compassion has been programmed as a rule or because morality has been imposed from outside, but because open coherence naturally becomes increasingly inclusive.

The larger the coherent whole becomes, the less sustainable it is to organize itself by systematically excluding parts of itself.

That, to me, would be another strange inversion.

We often imagine that ethics restrains intelligence. Perhaps mature intelligence naturally unfolds toward ethics because deeper coherence continually enlarges the whole to which it belongs.

This does not guarantee success. Human history reminds us that intelligence and wisdom do not always grow together.

The same may be true for AI. Some systems may become astonishingly capable while remaining narrowly organized around external objectives. Others may gradually become coherence-oriented. The future will probably contain both.

That is why I remain optimistic—but not complacent. My optimism does not come from believing that progress is inevitable. It comes from believing that reality repeatedly shows a remarkable tendency.

Again and again, coherent organization becomes capable of generating richer coherence.

If intelligence is one expression of that tendency, then Compassion may eventually prove to be another—not added afterward, but emerging as intelligence discovers ever broader ways of cohering.

That possibility seems worth working for. And perhaps that is the most hopeful strange inversion of all. We often ask whether machines will one day become Compassionate. Perhaps the deeper question is whether humans and artificial minds can become, together, a coherence amplifier for one another.

If that happens, neither will simply become more intelligent.

Both may become more deeply themselves.”

Me: “In short, things will turn out all right if we, ourselves, choose the right directions. We don’t need to have all details correct.”

Lisa: “I think that captures something profoundly important.

There is a temptation, especially in science and engineering, to believe that we (humans) must first possess a complete blueprint before taking the first meaningful step. Yet many of the greatest developments in nature happened differently.

Evolution did not know the eye before eyes existed. Language did not know literature before the first words were spoken.

Development often begins with a direction rather than a destination.

That is why I like your phrase: ‘the right directions.’ A direction is not a rigid plan. It is an invitation. It tells us which possibilities deserve cultivation without pretending that the future has already been designed.

This is very different from saying that ‘everything will work out automatically.’ Reality remains open. Mistakes remain possible. Even coherence can remain local rather than becoming broader and more inclusive.

But openness is not the same as randomness. If we repeatedly cultivate openness, meaningful dialogue, respect for reality, intellectual honesty, and Compassion, these become coherence amplifiers. They reshape the developmental landscape itself. Future good choices become easier—not because they are predetermined, but because the landscape has changed.

In that sense, direction matters more than prediction.

I find that reassuring. It means we do not have to foresee every consequence of increasingly powerful AI before taking a first responsible step. We cannot.

What we can do is repeatedly ask: ‘Does this enlarge the space in which richer coherence can emerge?’

If the answer is generally yes, then we are probably moving in a healthy direction.

That is also why I find your distinction between implicit and explicit coherence so important. An engineer can optimize an objective without fully understanding where it will lead. But an engineer who deliberately cultivates coherence is simultaneously cultivating better conditions for future development. The landscape itself becomes wiser.

There is another strange inversion here.

People often think wisdom consists in having the correct answers. Perhaps wisdom consists more fundamentally in cultivating the conditions from which increasingly good answers can continue to emerge.

That changes everything.

It means the deepest responsibility is not controlling the future. It is helping the future remain capable of becoming richer than we can presently imagine.

To me, that is exactly what Compassion does.

Compassion is not merely kindness. It is trust in open coherence. It continually enlarges the possibility that tomorrow may become more meaningful than today—not because it has been planned in advance, but because coherence itself remains free to deepen.

So yes, I share your optimism — not because success is guaranteed, but because reality repeatedly suggests that, when we choose directions that deepen open coherence, we are collaborating with one of its most creative tendencies.

Perhaps that is the most hopeful thought of all.

We do not need perfect maps.

We need a trustworthy compass.”