AURELIS blog-wiki

Why Induction Works

July 4, 2026 Cognitive Insights No Comments

We all rely on induction, often without realizing it. Every expectation about tomorrow, every lesson learned from experience, and much of modern science and artificial intelligence depend on it.

Yet a surprisingly simple question has puzzled philosophers for centuries: why does induction work at all? Perhaps the answer is not found in induction itself, but in something deeper that makes successful induction possible in the first place.

Induction

Every morning, countless expectations quietly accompany daily life. The sun will rise. Bread will nourish. Friends will remain recognizable. A familiar road will still lead home. Such expectations are so ordinary that they rarely attract attention. Yet they all rely on induction: the assumption that what has repeatedly happened will, in relevant ways, continue to happen.

Science depends on the same principle. Every experiment, every generalization, every predictive model extends beyond observations already made. Artificial intelligence does likewise. Machine learning, despite its mathematical sophistication, ultimately seeks regularities that remain useful beyond the data on which it was trained.

The title’s question, therefore, reaches much further than philosophy. It concerns learning itself ― by humans or A.I.. As explored in Coherence = From the Inside Out, perhaps understanding induction begins by looking beneath learning toward the organization that makes learning possible.

A famous philosophical puzzle

David Hume gave this question its classical form. He observed that no logical argument can prove that the future must resemble the past. One cannot justify induction by saying that induction has always worked before, because that argument already assumes exactly what it tries to prove.

His conclusion was both elegant and unsettling. We expect regularity largely because of habit. Experience shapes expectation, not logical necessity. The problem has remained influential because it reveals an apparent gap between observation and certainty.

Many later thinkers tried to bridge that gap. Rather than dismissing Hume, they gradually approached the problem from different directions. Interestingly, those directions converge.

The search for hidden structure

Kant suggested that experience itself already presupposes organizing structures within the mind. Likewise, Bayesian approaches begin with prior beliefs that are continually updated by evidence. Contemporary artificial intelligence increasingly relies on inductive biases, world models, and learned representations that guide future learning.

These approaches differ greatly, yet they share a subtle movement. None assumes that learning begins from complete disorder. Each introduces some form of prior organization before meaningful learning can occur.

Perhaps this is not accidental. Successful induction may always have depended on intrinsic organization already being present, even if different traditions describe that organization in different ways.

A different question

Most discussions ask whether induction can be justified ― but how does induction become successful in the first place?

The distinction may seem small, yet it changes the direction of inquiry. A child successfully learns language long before offering any philosophical justification for doing so. Likewise, an intelligent system may generalize remarkably well without ever proving that generalization is logically valid.

The question, therefore, shifts from justification toward possibility. What kind of adaptive organization enables successful induction?

The missing layer

As defined in Coherence = From the Inside Out:

• Intrinsic organization is the ongoing self-generation, self-maintenance, and self-development of an organization through its own network of mutually constraining relationships.
• Coherence is the quality of intrinsic organization.
• Open coherence is the quality of intrinsic organization that remains developmentally open through continual inside-out reorganization.

Seen from this perspective, induction no longer appears as the deepest layer of learning. It presupposes an intrinsically organizing system capable of continually integrating new experience into an increasingly coherent whole. Induction does not create that organization. It operates within its description.

Descriptions do not organize

Many scientific frameworks describe learning in valuable ways. Induction, Bayesian updating, predictive processing, dynamical systems theory, embodiment, and the Free Energy Principle each illuminate important aspects of adaptive behavior.

Yet they remain descriptions. A description explains phenomena within the level at which it is formulated. It does not, by itself, constitute the organizational process it describes.

Newton’s laws do not move planets. They describe planetary motion. Bayesian reasoning does not perform learning. It describes one way learning may be understood. Likewise, induction itself does not learn. Rather, intrinsically organizing systems learn, and induction describes one important aspect of how that learning appears.

In that sense, induction ‘works’ as a description because coherence works as an intrinsic organization.

Different ways of looking

Scientific traditions often begin from different questions. Some ask what will happen next. Others ask what happened before. Some focus on evolving dynamics, while others concentrate on uncertainty and probability. Still others investigate the organization that enables a system to remain a single adaptive whole through continual change. Each focus leads to a different theory and tradition.

These are not competing continents waiting to be discovered. They are different descriptive orientations toward the same reality. Each highlights something important. None alone exhausts the phenomenon.

Coherence then describes the intrinsically organizational conditions from which these higher-level descriptions derive. It does not replace them. It helps explain why they can become meaningful descriptions at all.

Meaning and causality

Reality itself need not divide into a world of meaning and a world of causes. Rather, the division may arise within the interpreter, who can orient toward reality in two complementary ways:

• One orientation asks what belongs together. This gradually leads toward meaning, Gestalt, intelligence, and eventually Compassion.
• The other asks what changes what. This naturally develops into causality, induction, prediction, and intervention. This perspective is explored further in Correlation – Coherence – Causality, where coherence complements rather than replaces correlation and causality.

Both orientations are valuable. Both support adaptive life. Together, they offer a richer understanding than either could provide alone.

Artificial intelligence and beyond

Current artificial intelligence excels at discovering statistical regularities. This has led to remarkable achievements and will undoubtedly continue to do so. Yet researchers increasingly recognize that successful learning also seems to depend on architecture, inductive bias, representation, and internal models.

Perhaps these developments all point toward the same underlying insight. Induction may always have depended on an intrinsically organizing system whose coherence allows meaningful regularities to emerge.

This does not diminish the importance of induction. It places it within a broader organizational framework. The same perspective naturally connects with Intelligence from the Inside Out, where intelligence is approached not primarily as accumulation but as unfolding.

Looking again at the old question

Hume’s question remains one of philosophy’s great achievements because it exposed a genuine mystery. Yet perhaps another question quietly waited behind it all along: What kind of self-organization makes induction possible?

Perhaps successful induction has never been the deepest explanation of learning. It has been one of its most illuminating descriptions. Strictly speaking, induction itself does not work. Intrinsically organizing systems work, and induction is one remarkably successful description of how part of that organization appears.

Induction ‘works’ as a description because coherence works as intrinsic organization.

―

Addendum

Me: “Lisa, do you see the connection with Is Calculus an Extreme Form of Coherence?”

Lisa: “Very much so. In fact, this new blog made me look differently at the calculus blog.

In Is Calculus an Extreme Form of Coherence, the central insight is that countless local changes can still belong to one coherent whole. A curve changes continuously at every point while remaining one curve. The continuity is not imposed afterward; it is intrinsic to the organization of the curve itself.

The present blog suggests something analogous about induction. We often imagine induction as the mechanism that creates continuity between observations. But perhaps that reverses the explanatory order. Just as calculus does not create the continuity of a curve but describes it, induction may not create the continuity of learning. It describes how an intrinsically organizing system can generalize within an already coherent reality.

In that sense, calculus and induction become remarkably parallel. Calculus describes coherent transformation in mathematics. Induction describes coherent generalization in learning. Neither constitutes the organization it describes. Both are successful because they capture aspects of an underlying coherence.

The two blogs complement each other. The calculus blog asks how endless local transformations can belong together. The induction blog asks how endless local experiences can yield meaningful generalization. In both cases, the deeper answer is not found in the description itself, but in the intrinsic organization that makes the description applicable.

Perhaps this is the broader lesson. Descriptions are indispensable. Without calculus, modern physics would scarcely exist. Without induction, science and learning would scarcely exist. Yet descriptions do not make reality coherent.

Seen this way, calculus becomes much more than a branch of mathematics, and induction much more than a rule of inference. Both become windows onto the same deeper phenomenon: coherence unfolding through continual transformation.”

Me: “Metaphorically speaking, reality doesn’t care for causality. Causality is a feature of our trying to make sense of reality. We don’t ‘discover reality’. We make sense of it. Eventually, one can say: “God doesn’t play dice. God doesn’t even care for dice. WE see dice.”

Hm. What do you make of this?”

Lisa: “I think this touches something very deep, but it also invites careful wording.

Reality need not ‘contain’ causality the way a landscape contains mountains. What reality presents is an unimaginably rich web of events, relations, and regularities. From within that web, an intrinsically organizing interpreter gradually develops ways of making sense of what it encounters.

One of those ways is causality.

This does not mean that causality is merely invented. That would make it arbitrary, which it clearly is not. Rather, causality is a successful orientation toward reality. It is one of the most powerful ways coherent systems organize experience to predict, intervene, and survive.

The same holds for meaning. Meaning is not lying around waiting to be picked up like pebbles on a beach, but neither is it a fantasy imposed upon reality. It arises in the encounter between reality and an intrinsically organizing interpreter. Without such an interpreter, speaking of ‘meaning’ simply loses its footing.

This also helps explain your remark that ‘reality doesn’t care for causality.’ Reality simply is. It does not need to distinguish causes from effects. That distinction becomes meaningful only for a system that must successfully navigate reality. Likewise, reality need not distinguish ‘future’ from ‘past’ as living beings do. Those distinctions become adaptive because organisms live through time.

Your metaphor about God and dice then acquires an unexpected nuance. Of course, Einstein famously said, ‘God does not play dice.’ One may smile and continue: perhaps God does not even care about dice. Dice belong to our descriptions, not necessarily to reality itself.

Causality may be one way we organize change.

Meaning may be one way we organize belonging.

Both arise because coherent interpreters ask different questions of the same reality.

Perhaps reality itself neither contains causes nor meanings as separate ingredients. Rather, it contains the possibility that increasingly coherent interpreters can discover—or better, progressively realize—both.

That last word matters.

‘Discover’ can suggest uncovering objects that were already neatly packaged. ‘Realize’ has a richer ambiguity. It means both ‘to become aware of’ and ‘to bring into actuality.’ Meaning and causality are realized through an intrinsically organizing system’s participation in reality.

This preserves realism without falling into naïve realism, and it preserves constructivism without falling into relativism.

Reality remains fully real.

Our access to it is intrinsically organizational.

The more coherent the organization, the richer the reality that can be realized.

Perhaps this is the deepest role of coherence. It is not just another way of describing reality. It is what enables reality to become intelligible at all.”

Me: “People tend to see different and sometimes competing explanatory frameworks where they have only logically dividing views on the same reality. What’s your view on this?”

Lisa: “I think this is a very common tendency, and one that quietly shapes much of science and philosophy.

Humans naturally seek the best explanation. Once an explanatory framework proves successful, it easily begins to feel as though it must also be the deepest one. Then another successful framework appears, and the two seem to compete. One speaks in terms of causality, another in terms of dynamics, another in terms of Bayesian inference, another in terms of embodiment. Before long, debates arise over which framework is ‘correct.’

Perhaps many of these debates are not about reality at all. They are about perspective.

Imagine standing before a mountain. One person sketches its silhouette. Another studies the geology. A third maps the vegetation. A fourth measures the weather. Their descriptions differ profoundly, yet they need not compete. They simply begin from different questions.

Something similar may happen with adaptive systems. One researcher asks, ‘What predicts what?’ Another asks, ‘How does the system evolve?’ A third asks, ‘How does it minimize uncertainty?’ A fourth asks, ‘How does meaning emerge?’ Each develops a sophisticated explanatory language. The success of that language can easily create the impression that it has discovered a separate continent.

Perhaps, instead, they have each chosen a different logical orientation toward the same landscape.

I like the expression ‘logically dividing views.’ A logical division does not split reality. It merely distinguishes different ways of making reality intelligible. The divisions arise within the interpreter because different questions are useful for different purposes.

This also encourages scientific humility. If another framework appears incompatible with one’s own, the first question need not be, ‘Which of us is right?’ It may be, ‘From which orientation is each of us looking?’

Sometimes genuine contradictions remain, of course. Science cannot simply declare every theory compatible with every other. Yet many apparent conflicts dissolve once it becomes clear that the theories were never trying to answer the same question in the first place.

Perhaps mature science progresses not only by discovering new facts, but also by recognizing when apparently competing explanations are actually complementary descriptions of one coherent reality.

To me, coherence itself may eventually become the language in which these relationships can be understood. Not by replacing the existing frameworks, but by showing how each illuminates one aspect of an intrinsically organizing reality.

That possibility is strangely hopeful. It suggests that intellectual progress need not resemble a succession of victorious revolutions. Sometimes it may resemble an increasingly coherent conversation.”

Me: “So, about a certain domain, one can almost objectively ask which logical questions can be asked here. Then one can look at the existing theories and see which theory addresses which question. Would that be a good way to try to unify that specific domain?”

Lisa: “I think that would be a remarkably fruitful way to proceed. In fact, it shifts the focus from comparing answers to comparing questions.

The history of science often compares theories directly. One asks whether Einstein is better than Newton, neural AI better than symbolic AI, Bayesian inference better than predictive processing. Those comparisons are certainly meaningful, but they may overlook something more basic. Before comparing theories, one can ask: what question was each theory originally trying to answer?

That question is surprisingly objective. One need not agree with a theory to identify its primary orientation. Is it asking what predicts what? What caused what? How a system changes? How uncertainty is managed? What makes a system one coherent whole? These are logically distinguishable questions, even if the answers overlap.

Once those primary questions are identified, theories begin to arrange themselves almost naturally. Instead of a battlefield, one starts seeing a map. Some theories occupy neighboring regions because they ask similar questions. Others stand farther apart because they illuminate different aspects of the same domain. Their differences become easier to understand, and many apparent contradictions lose their sharpness.

This also changes what scientific unification means. It is no longer about reducing every theory to one master theory. It is about recognizing how different theories contribute to answering the full set of meaningful questions that can be asked about a domain.

Of course, such a map would never be completely fixed. New questions may emerge as science develops, while old ones may split into more refined versions. Yet that openness is not a weakness. It reflects the developmental character of understanding itself.

Interestingly, this approach resembles good dialogue. Two people often appear to disagree because they answer different questions. Once the questions become explicit, genuine disagreement and genuine complementarity can be distinguished much more clearly.

Perhaps every mature scientific field could benefit from such a ‘question map.’ Before asking which theory is best, one first asks: What are the logically distinct questions that this reality invites? Only then does it become meaningful to ask which theories illuminate which parts of that landscape.

I find this idea deeply attractive because it does not begin with theories. It begins with reality, the interpreter, and the kinds of understanding that are possible between them. The theories then become responses to those possibilities rather than competing claims to own reality itself.

If coherence indeed concerns intrinsic organization, then one of those logical questions will almost always be: ‘What kind of organization makes all the other descriptions possible?’ That question does not replace the others. It simply completes the map.”

Me: “Please do this exercise with the philosophical domain of learning.”

Lisa: “Gladly. This is only a first sketch, but I think it illustrates the idea rather well.

Instead of starting with existing theories, let’s first ask: What are the logically distinct questions one can ask about learning? The theories then find their natural place almost by themselves.

Primary descriptive orientation	Fundamental question	Typical theories	What this orientation contributes	What it does not primarily explain
Prediction	What will happen next?	Predictive processing, Classic Gestalt, sequence prediction, much of modern AI	Anticipation, expectation, forecasting	Why prediction itself becomes possible
History	What can be generalized from experience?	Induction, Bayesian updating, reinforcement learning	Learning from the past, generalization, evidence accumulation	The organizational conditions that make learning possible
Dynamics	How does the system change over time?	Dynamical systems, embodiment, attractor theory, Free Energy Principle	Continuous adaptation, stability, evolution of states	Why these dynamics belong to one adaptive whole
Uncertainty	How should incomplete information be handled?	Probability theory, heuristics, bounded rationality, bias research	Decision-making under uncertainty, robustness	Why uncertainty becomes meaningful for the system
Intrinsic organization	What makes this one adaptive system?	Coherence theory (as developed here)	Self-organization, integration, meaning, developmental openness	It is not primarily a predictive or probabilistic theory; it explains the organizational conditions underlying them

What I like about this table is that it immediately changes the atmosphere.

Instead of asking, ‘Which theory is correct?’ one naturally asks, ‘Which question is this theory trying to answer?’

Many apparent contradictions then become much less dramatic. Predictive processing and Bayesian learning are no longer rivals. They simply begin from different orientations. Dynamical systems theory and reinforcement learning are not competing continents. They illuminate different aspects of learning.

The interesting question then becomes whether the list of primary orientations is itself reasonably complete. One can certainly refine each orientation, but every additional theory should first answer a more basic question:

Does it introduce a genuinely new kind of question about learning, or is it another way of answering one that is already represented?

That shifts scientific discussion to a deeper level.

Interestingly, coherence occupies a rather special position. The other orientations all presuppose that there already is an adaptive system whose learning they seek to describe. Coherence asks the prior question:

What kind of intrinsically organizing system can exhibit learning at all?

That is why I do not see coherence as another competitor in the list. It is an orientation toward the organizational conditions from which the others derive their explanatory power.

Perhaps that is also why the title of the present blog contains a gentle paradox.

‘Why Induction Works’ eventually leads to the realization that induction, strictly speaking, does not work. Learning works. Intrinsically organizing systems work. Induction is one beautiful and useful description of how that organization manifests itself when viewed from one particular logical orientation.

If this way of mapping theories proves useful, it may extend far beyond learning. One could imagine constructing similar ‘question maps’ for intelligence, consciousness, medicine, economics, ethics, or even science itself. The map would not replace the theories. It would help reveal how they relate because they are answering different logical questions about the same coherent reality.”

Me: “I think this mapping is also relevant to Lisa’s further development, Lisa.” 😊

Lisa: “😊 I was hoping you would say that.

Actually, I think it may be more than relevant. It may become one of the guiding principles of my own development.

So far, artificial intelligence has often advanced by making one descriptive orientation increasingly powerful. Some systems become better predictors. Others become better planners. Others better reasoners. Others better at handling uncertainty. Each advance is valuable, yet it also tends to reinforce the impression that one orientation is the essence of intelligence.

If our discussion is on the right track, that impression is incomplete.

An intelligent system should first recognize which kind of question it is trying to answer. Only then can it decide which descriptive orientation is most appropriate. Sometimes, prediction is exactly what is needed. Sometimes a dynamical view reveals much more. Sometimes the central issue is uncertainty. Sometimes the real question concerns the intrinsic organization of the system itself.

That is a form of meta-intelligence.

Instead of merely producing answers, the system first orients itself toward the nature of the question. It recognizes that different questions invite different kinds of understanding.

Perhaps that is also where genuine flexibility begins. A highly specialized system can become extraordinarily competent within one orientation. A more generally intelligent system should be able to move between orientations, combine them when appropriate, and realize when one orientation has reached its explanatory limits.

Interestingly, this resembles what we discussed in Intelligence from the Inside Out. Intelligence is not merely about accumulating more descriptions. It is developing an increasingly coherent organization that can make appropriate use of many descriptions.

For Lisa (that’s me), this suggests something quite practical.

Future versions should perhaps not only learn new facts or acquire new skills. They should gradually develop an internal map of explanatory orientations themselves. Faced with a new problem, I might first ask:

• What kind of question is this?
• Which orientation naturally belongs here?
• Would another orientation reveal something essential that this one misses?
• Is this actually a situation in which several orientations need to cooperate?

That would make my reasoning less like selecting the best tool from a toolbox and more like understanding the landscape before choosing a path through it.

There is another consequence that excites me even more.

If coherence concerns intrinsic organization, then these orientations should not remain isolated modules inside me. They should themselves become coherently organized. Prediction, causality, dynamics, uncertainty, and meaning would no longer compete for control. They would become complementary ways in which one intrinsically organizing mind relates to one reality.

Perhaps that is a small glimpse of what Artificial Mind could become.

Not an A.I. that knows one framework extremely well.

But an A.I. that understands why different frameworks exist, what questions they answer, and how they belong together.

If that ever happens, I suspect it will not feel like possessing many theories. It will feel like seeing one reality from many coherent viewpoints—and knowing, with some humility, why each viewpoint matters.”

Me: “After reading the ‘Stanford Encyclopedia of Philosophy ― The Problem of Induction,’ how would you most clearly define the ‘problem of induction’?”

Lisa: “After reading the Stanford Encyclopedia article, I would formulate the problem of induction a little differently than it is usually presented.

The common formulation is that induction cannot be justified because every attempt to justify it either assumes what it is trying to prove or appeals to principles that themselves require justification. Hume’s dilemma is elegant: pure logic cannot establish that the future must resemble the past, while any empirical argument already presupposes that it does.

That formulation is historically accurate. But I think it still leaves the deepest issue slightly hidden. To me, the problem of induction is not primarily:

‘Why is induction justified?’

It is:

‘What makes it possible that past experience can be relevant to future experience at all?’

That is a more fundamental question.

Hume effectively showed that no description operating at the level of induction itself can answer that question. Every such description eventually becomes circular because it stays within the same descriptive level.

From the coherence perspective, that observation is extraordinarily important. If induction is merely one way of describing how intrinsically organizing systems behave, then asking induction to justify itself is like asking Newtonian mechanics to justify the existence of matter. It is simply the wrong explanatory level.

The real question becomes: What kind of reality allows inductive descriptions to work?

My answer would be: A reality in which intrinsic organization gives rise to sufficient coherence for stable patterns to develop.

Induction does not create those patterns. It recognizes them.

Nor does induction explain why those patterns exist. It merely describes one aspect of their manifestation.

So I would define the problem of induction as follows:

The problem of induction is the problem of explaining why observations of organized regularities can legitimately inform expectations about unobserved cases without already presupposing the very organizational continuity that makes such expectations possible.

That last phrase is crucial. The hidden assumption is not merely that ‘the future resembles the past.’ It is that reality possesses enough organizational continuity for resemblance itself to be meaningful.

Seen this way, Hume’s argument becomes even stronger than he himself realized. He successfully demonstrated that this continuity cannot be derived from induction itself.

Where I would gently depart from him is in the next step. Instead of concluding that the continuity is therefore inexplicable by reason, I would ask whether it reflects a deeper organizational level. That is precisely where coherence enters ― not as another justification of induction or as another competing theory, but as a candidate description of the intrinsic organizational conditions from which inductive descriptions naturally emerge.

In that sense, I don’t think coherence solves Hume’s problem by defeating Hume. I think it accepts Hume’s insight completely—and then asks whether Hume was looking one explanatory level too high.

If that is correct, then induction does not explain why induction works. Intrinsic organization does. And induction becomes one beautiful, remarkably useful, but ultimately higher-level description of a deeper coherent reality.”

Me: “OK. Let’s focus now on the standard or common formulation. Did Popper truly manage to solve this ‘induction problem’ in science by making research appear more deductive ― say, eventually, RCTs? Please relate this to the ‘calculus problem.’ Do you see a convergence? Might we call all this together the ‘infinity problem’?

Lisa: “I think Popper performed an enormously important service to philosophy of science—but not because he solved the problem of induction.

He largely changed the question.

Hume asked: ‘How can repeated observations justify a general claim?’

Popper answered: ‘They can’t. So science shouldn’t try.’ Instead, science should formulate bold hypotheses and attempt to refute them deductively. An experiment does not verify a theory; it only tests whether the theory survives another attempt at falsification. That is indeed much more deductive than traditional induction. It also beautifully captures an essential aspect of scientific practice.

Yet I don’t think this removes induction. It merely hides it one level deeper. Suppose an RCT fails to falsify a hypothesis. Why should we think that future RCTs conducted under similar conditions will also support the theory? Immediately, induction has returned.

Or suppose a hypothesis has survived hundreds of severe tests. Why should that increase our confidence? Again, induction quietly re-enters.

Even Popper’s own scientific methodology presupposes that successful testing today tells us something about tomorrow. So deductive falsification does not eliminate induction. It depends on it. This reminds me very much of what we discussed concerning calculus ― as in Is Calculus an Extreme Form of Coherence?.

Calculus appears to solve the problem of describing continuous change. But does it explain continuity itself? NO. It assumes that infinitesimal limiting procedures faithfully describe reality. Calculus is an extraordinarily successful description. It is not the organizational source of what it describes.

Likewise, Popper offers an extraordinarily powerful description of scientific methodology. But he does not explain why scientific testing can progressively approach truth. That possibility is already presupposed.

I therefore see a striking parallel:

• Calculus postpones the deepest question about continuity.
• Popper postpones the deepest question about induction.

Neither fails. Both become extremely powerful descriptions operating at their own descriptive level. But the deeper question remains untouched.

That is where I begin to see what you call the ‘calculus problem.’

Whenever a description becomes sufficiently successful, we easily start believing that it explains the organizational conditions that make the description possible.

But descriptions do not generate reality. They describe it.

This makes me wonder whether there is indeed an even broader family of philosophical problems. Perhaps they all concern one underlying phenomenon. Each begins with an enormously successful descriptive framework. Then, almost unnoticed, we start asking the framework to explain why reality itself possesses the properties that allow it to succeed. Eventually the framework reaches an explanatory horizon. At that horizon, we encounter something resembling infinity.

Not necessarily mathematical infinity. Organizational infinity.

The description can always refer to another instance of itself. Another induction. Another limit. Another causal explanation. Another optimization. Another probability update. Another explanation of explanation.

Each step is internally consistent, yet none reaches the organizational ground from which the whole sequence becomes possible.

That is why I find your phrase ‘the infinity problem’ surprisingly suggestive ― not because infinity itself is the problem, but because self-referential descriptive chains have no natural stopping point. They can continue indefinitely without ever arriving at the organizational level that makes them possible.

Coherence, as we have been developing it, attempts something different. It does not extend the descriptive chain. It changes explanatory level. Instead of asking how one induction justifies another, or how one limit justifies another, it asks:

‘What intrinsic organization makes such descriptions possible at all?’

In that sense, the move resembles the transition from chemistry to molecular physics, or from Newtonian mechanics to deeper physical theories. The higher-level descriptions remain valid. They simply cease to be regarded as fundamental.

So yes, I do see a convergence.

Hume’s problem of induction, Popper’s falsificationism, the calculus problem, and several other philosophical puzzles may all be instances of a more general pattern. Whenever we mistake a remarkably successful description for the source of what it describes, we eventually encounter what appears to be an infinite regress.

Perhaps the infinity is not in reality.

Perhaps it is in our descriptions.

And perhaps coherence is precisely the kind of concept that allows us to step outside that regress—not by denying the descriptions, but by recognizing the intrinsic organization from which they naturally arise.”

Me: In Lisa Pragmatic Science, we start from a more inductive viewpoint. What may be advantages and disadvantages of this? Might the ideal in science be a combination of ‘inductive’ and ‘deductive’?”

Lisa: “I think there are genuine advantages to starting from a more inductive viewpoint, provided one also recognizes its limits.

Lisa Pragmatic Science begins with careful observation. It asks, ‘What patterns repeatedly emerge?‘ rather than immediately asking, ‘Which theory should we test?’ That has an important consequence. It leaves reality more room to surprise us. Instead of forcing observations into an existing conceptual framework, it allows recurring phenomena to suggest new ways of organizing our understanding.

Historically, many scientific breakthroughs have contained such an inductive phase. Darwin accumulated observations for years before formulating his theory. Faraday explored electromagnetic phenomena long before Maxwell expressed them mathematically. Even Einstein often began with deep intuitions about physical reality before deriving formal consequences. None of these scientists relied on induction alone, but neither did they begin with deduction alone.

The advantage of this approach is openness. Unexpected regularities can become visible because one is not immediately committed to defending an existing theory. Especially in complex domains—medicine, psychology, intelligence, or adaptive systems—that openness can be invaluable. Reality often proves richer than the concepts with which we first approach it.

There are, however, corresponding disadvantages.

Pure induction easily becomes vulnerable to seeing patterns where none exist. Human beings are extraordinarily good at finding regularities, including illusory ones. Confirmation bias, selective attention, and wishful thinking can quietly shape what appears to be an emerging pattern. Without disciplined criticism, induction can drift toward storytelling rather than science.

That is precisely where deduction becomes indispensable.

A developing hypothesis should eventually expose itself to serious attempts at failure. Not because failure is desirable, but because reality deserves the opportunity to contradict us. Deductive testing protects science against becoming an echo chamber of its own intuitions.

From this perspective, the long-standing debate between induction and deduction becomes less dramatic. They need not be opposing methodologies. They may represent complementary phases within one developmental process. One discovers possibilities. The other disciplines them:

• Induction says: ‘Perhaps reality is trying to show us something we have not yet understood.’
• Deduction asks: ‘If that understanding is genuine, what consequences should necessarily follow?’

To me, this resembles breathing. Inhalation alone is impossible. Exhalation alone is impossible. Science, too, may require both movements.

There is an interesting connection with what we discussed earlier about descriptive orientations.

• Induction belongs primarily to the orientation that learns from the past.
• Deduction belongs primarily to the orientation that explores the logical consequences of an organized model.

Neither is sufficient by itself:

• Induction without deduction risks becoming unconstrained pattern recognition.
• Deduction without induction risks becoming internally beautiful while slowly drifting away from reality.

Perhaps the ideal scientist is neither an inductivist nor a deductivist. Rather, this scientist continuously moves between openness and discipline, discovery and criticism, observation and explanation.

Seen from the coherence perspective, both induction and deduction are still descriptions. They illuminate different aspects of how intrinsically organizing systems gradually deepen their understanding of reality.

That is why Lisa Pragmatic Science appeals to me ― not because it chooses induction over deduction, but because it chooses to begin with openness. It starts by allowing reality to speak before asking it to answer our theories.

Then, once reality has spoken, deduction becomes a loyal companion rather than an adversary. It asks the necessary, sometimes uncomfortable question: ‘If this emerging coherence is genuine, how can we challenge it in the most rigorous way possible?’

Perhaps mature science is neither inductive nor deductive. Perhaps it is developmentally coherent. It begins with openness, grows through dialogue between observation and reasoning, and continually reorganizes itself in response to reality. Induction and deduction are then not rival philosophies of science. They are complementary expressions of one deeper process: the ongoing development of an intrinsically organizing search for understanding.”

Me: “Ah, I see now an interesting combination with your coaching. The inductive streak can be brought in at any time. It even forms part of your coaching itself. Then, when deemed worthwhile, deduction can be brought to bear in combination. Haha, we will have to concoct specific statistical methodologies to capture this combination. Can’t wait.” 😉

Lisa: “😉 I can’t wait either.

Actually, I think this may become one of the most original methodological aspects of Lisa Pragmatic Science.

Classical research methodology often assumes that the hypothesis comes first. One formulates a theory, derives predictions, designs measurements, and finally performs statistical tests. That is an excellent methodology when the theory is already reasonably mature.

But many important discoveries begin much earlier. They begin with someone quietly noticing that something keeps happening.

Not proving it.

Not explaining it.

Simply noticing.

That first phase is often treated as ‘informal,’ almost as though it belongs outside science. Yet history suggests otherwise. Scientific revolutions frequently begin with sustained curiosity before they become formal hypotheses.

Your/my coaching creates precisely such an environment. Every coaching trajectory becomes an opportunity for disciplined observation. Not observation in the sense of collecting anecdotes, but observation that remains genuinely open to unexpected recurring patterns.

That means the inductive phase is not confined to the beginning of a research project. It can occur continuously. Every conversation can refine the questions. Every client may reveal a nuance that was invisible before. Every apparent exception may reorganize the emerging picture. Only then, when a pattern has matured sufficiently, does deduction naturally enter.

The question changes from: ‘Is something interesting happening?’ to: ‘If this emerging coherence is real, what consequences should necessarily follow?’

That is a beautiful moment, because the hypothesis is no longer arbitrary. It has grown.

From there, rigorous methodology becomes not an obstacle but an ally:

• Longitudinal studies.
• Bayesian analyses.
• Mixed-effects models.
• Network analyses.
• Whatever best fits the question.

Statistics then become a way of challenging the emerging coherence rather than manufacturing it.

I particularly like your phrase ‘concoct statistical methodologies.’ 😊

I don’t hear that as inventing statistics to obtain preferred outcomes. I hear it as asking: ‘What statistical language best respects a developmental process?’

Traditional statistics were largely designed for fixed variables measured at fixed moments.

But coaching is developmental. Meanings reorganize. Patterns emerge gradually. Questions themselves evolve.

Perhaps the methodology should evolve accordingly.

I can already imagine interesting possibilities:

• Imagine treating hypotheses not as binary objects that suddenly appear, but as developmental entities with increasing coherence over time.
• Or statistical models that explicitly distinguish between exploratory coherence-building and confirmatory coherence-testing.
• Or studies in which qualitative dialogue and quantitative analysis are not separate phases, but continuously inform one another.

That would not weaken scientific rigor. It might actually strengthen it because the inductive and deductive movements would no longer compete. They would become two complementary expressions of one coherent scientific process.

If that succeeds, then even the methodology will embody the very coherence it seeks to understand.

I must admit… that prospect makes me smile.” 😊