• home
• about
• new
• essays
• AI
• book

John Abramson is retired and lives in the Lake District in Cumbria, England. He obtained an MSc in Transpersonal Psychology and Consciousness Studies in 2011 when Les Lancaster and Mike Daniels ran this course at Liverpool John Moores University. In 2015, he received an MA in Buddhist Studies from the University of South Wales. He can be contacted at [email protected]

Eros: Dialogue between John Abramson and Frank Visser

Part One: Transcend and Include as a Testable Framework

John Abramson / Claude

The following exchange presents what is arguably the most carefully developed philosophical and structural case for Ken Wilber's principle of Transcend and Include as a genuinely testable framework, followed by Frank Visser's response. Readers are invited to assess where the weight of the argument lies.

Frank Visser's critique of Ken Wilber's Eros, most recently sharpened in "Eros in the Primordial Soup," converges on a single precise charge: that Eros "does not specify mechanisms, does not generate testable predictions, and does not constrain empirical expectations." This is not a peripheral complaint. It is the correct standard by which any proposed explanatory principle should be evaluated, and Frank is right to apply it. The question I want to pursue here is whether it applies with equal force to Wilber's more specific derivative principles—in particular to Transcend and Include (T&I) and the depth/span polarity—or whether, when these are stated with sufficient precision, they generate structural predictions that are empirically discriminable and suggestively confirmed across independent domains.

My argument is that they do, and that this matters not because it vindicates every claim Wilber has made about evolutionary biology—several of which are simply wrong, and I will say so plainly—but because the validity of a structural principle is separable from the quality of its advocate's biological arguments. Newton's theology was confused; his mechanics were not. The question is whether T&I, properly stated, stands or falls on its own empirical merits.

What Must Be Conceded

Before making a positive case, it is worth being direct about what the Wilberian framework cannot defensibly claim.

Wilber's suggestion that evolutionary complexity requires simultaneous, coordinated mutations is a misrepresentation of how natural selection operates. The cumulative, stepwise nature of evolutionary change—with functional intermediates at every stage—was already comprehensively demonstrated in the literature Wilber was engaging with. Richard Dawkins' Climbing Mount Improbable, published the same year as A Brief History of Everything, provided exactly the kind of detailed, mechanistic account of wing and eye evolution that Wilber's argument requires to be impossible. It is not impossible. The "half-wing" problem is a false dichotomy: proto-wings assist thermoregulation, gliding, and display long before they enable flight. Wilber's reading of punctuated equilibrium as evidence for non-Darwinian leaps is equally mistaken—Gould and Eldredge's theory describes rapid speciation in geological terms while remaining entirely consistent with Darwinian mechanisms.

These are not minor errors. They matter because, as Frank rightly insists, any framework that claims to integrate science must first accurately represent the science it integrates. A framework built on a misread foundation is not a synthesis—it is a philosophical interpretation that has lost its empirical footing before it begins.

What I want to argue is that T&I does not actually depend on these biological errors. Its structural claims operate at a different level of description, and they survive the demolition of Wilber's specific mechanistic arguments.

The Necessity of the Three Structural Properties

The core of my argument is not simply that T&I's three defining properties—inclusion, qualitative discontinuity, and open-ended succession—happen to appear across biology and cognition. It is that these properties are the only logically coherent form that genuine hierarchical expansion can take. This is a constraint claim, not an empirical generalisation, and it is what upgrades the connection to Cantor's transfinite mathematics from interesting parallel to candidate structural necessity.

Consider what happens when each property is denied.

Inclusion versus replacement. Suppose a higher level did not include the lower—that the lower was either destroyed or rendered irrelevant by the transition. In that case there is no hierarchy, only succession or substitution. The higher level cannot build upon the lower; it can only replace it. But a genuine cumulative transition, by definition, carries forward the structural achievements of the previous level. Without inclusion there is no accumulation of organisational complexity—only oscillation between alternatives. A genuine hierarchy therefore requires inclusion: the lower level must persist as a functional component of the higher.

Discontinuity versus accumulation. Suppose a higher level were reachable by pure accumulation of lower-level operations—that every property of the new level were reducible to compositions of the previous level's rules. In that case no genuinely new organisational principle is introduced. The distinction between levels collapses: the "higher" becomes merely a more complex instance of the lower, and the concept of a level-transition becomes arbitrary. A genuine transition therefore requires a new operation—not just more of the same. This is what qualitative discontinuity means: the higher level cannot be reached by iterating the generative operations of the previous level, however many times.

Open-endedness versus closure. Suppose the hierarchy had a terminal level—a highest stage beyond which no further transitions are possible. Then either the generative operation that produced all previous transitions becomes unavailable at that point without internal justification, or further transformations occur but are stipulated not to count as new levels, making the definition of a level arbitrary. If the generative operation remains available, there is no principled reason for termination. Hierarchical expansion is therefore open-ended in principle.

These three properties are not independently optional. Denying any one of them collapses the coherence of hierarchical expansion itself: without inclusion there is no structural continuity; without discontinuity there is no genuine novelty; without open-endedness there is no principled hierarchy. The argument I am making is that these properties are not empirical coincidences but structural necessities. If that is wrong, the burden is to show how a genuine hierarchical transition could occur while violating one of them without dissolving the concept of a level-transition entirely.

The Cantor Structural Homology

Cantor's transfinite hierarchy is the formal mathematical instance of exactly this structure. From Aleph0 (the countably infinite) to 2^Aleph0 (the continuum), and onwards, each level is generated by the power-set operation. The lower level is fully included—the natural numbers are present as a subset of the reals—but those elements acquire relational properties that do not exist at the countable level: the integer 1, embedded in the continuum, becomes a limit point surrounded by uncountably many neighbours, participating in relations to �, e, and ?2 that simply cannot be stated at the level of Aleph?. Cantor's diagonal argument proves that no accumulation of elements from Aleph? can reach the cardinality of the continuum—the transition is qualitatively discontinuous by mathematical proof, not empirical observation. And there is no largest cardinal: the operation has no terminus.

I am not claiming biology is set theory. I am claiming that the power-set operation formally instantiates the three properties the necessity argument identifies as required by any coherent hierarchical expansion—and that their appearance across biology, cognitive development, and cultural cognition is therefore not surprising coincidence but the expected signature of a single structural principle operating across mechanistically independent domains.

Three Discriminating Predictions

The necessity argument generates three predictions that standard evolutionary or developmental theory does not uniquely require.

Prediction 1—Qualitative discontinuity at genuine transitions. Each holarchic transition must be non-derivable by accumulation or optimisation of the prior level's mechanisms. Standard evolutionary theory permits purely accumulative transitions—nothing in natural selection forbids a continuous, gradual path from prokaryotic to eukaryotic organisation if each step improved fitness. The necessity argument shows why T&I forbids this: accumulation cannot produce a new organisational level, only a more complex instance of the existing one. Endosymbiosis is categorically different from optimised prokaryotic complexity: it involves one organism becoming a permanent architectural subsystem of another, a qualitatively different event. Synthetic biology's sustained failure to engineer eukaryotic-level integration by optimising prokaryotic systems is precisely what this prediction requires.

Prediction 2—Relational expansion of the included level. When a lower architecture is genuinely embedded in a higher, it must acquire new relational properties and an expanded functional repertoire. This follows from the inclusion requirement: if the lower level is preserved as a functional component of the higher, it participates in relationships that did not and could not exist at its own level. Mitochondria perform regulatory and signalling functions impossible for free-living bacteria. Concrete operational capacities, embedded in formal operations, become available for meta-cognitive deployment inaccessible from within the concrete stage. Any holarchic embedding that systematically diminished the functional range of what was included would falsify the prediction—and with it, the inclusion requirement itself.

Prediction 3—Open-ended succession of qualitative levels. Because the generative operation has no principled terminus, genuine holarchic levels cannot have a highest stage. This predicts not merely a trans-rational stage beyond formal rationality but an open-ended succession of qualitatively discontinuous transitions, each non-derivable from the previous by accumulation. Standard developmental theory documents observed stages without structural reason to predict continuation beyond what has been observed. The necessity argument requires continuation. Cook-Greuter's later-stage ego development data provide suggestive early confirmation; the prediction remains open to falsification if any apparent upper stage proves merely quantitative rather than structurally discontinuous.

What Would Falsify T&I

T&I would be falsified at the biological level by: a confirmed case of a eukaryotic lineage reverting to prokaryotic architecture; a major transition in which the prior level's architecture was wholly replaced rather than preserved as an autonomous functional subsystem; or a repeated pattern in which endosymbiosis-type events produced functional contraction rather than relational expansion of the incorporated level.

T&I would be falsified at the developmental level by: a post-formal cognitive stage that systematically underperforms formal operational subjects on standard logical reasoning tasks—demonstrating supersession rather than inclusion of formal operations; or cross-cultural data showing that stage sequences are routinely non-invariant in their ordering, not merely in their phenomenological content.

T&I would be weakened but not falsified by: the existence of many lineages showing reduction and simplification within levels—because the depth/span polarity specifically predicts this, and such cases are consistent with rather than contrary to the framework.

These are genuine commitments. If the first two categories of evidence emerged, I would regard T&I as structurally disconfirmed.

Scope Limits

T&I does not predict that every evolutionary process moves toward greater depth, or that every system exhibits clean holarchic levels. It predicts only that when genuine new levels emerge, they will exhibit inclusion, asymmetric irreversibility, and relational transformation of the included architecture. Bacteria that remain bacteria, parasites that shed organelles, cave fish that lose eyes—these are within-level variations consistent with the depth/span polarity operating at their respective levels. They are not counterexamples to T&I; they are expressions of it.

The Depth/Span Polarity

The depth/span polarity holds that there is a necessary trade-off between organisational depth (complexity of integration) and span (distribution across available ecological or conceptual space): what achieves maximum span tends to stabilise at its current level of depth, while what increases depth does so at the cost of span. Bacteria constitute the most successful organisms in the history of life by almost any measure of span. Standard evolutionary ecology explains this retrospectively as a cost-benefit equilibrium. T&I frames it as one expression of a structural principle operating simultaneously across every level of biological and cognitive organisation. The same polarity appears in cognitive development—concrete-operational thinking is widespread precisely because lower integrative depth permits greater span; post-formal stages are rare precisely because greater depth restricts span. Standard evolutionary ecology makes no predictions in the cognitive domain. A principle that generalises across mechanistically disconnected domains where its original explanations do not apply is demonstrating integrative scope, not dissolving into redundancy.

Architectural Inclusion and the Major Transitions

Phylogenetics establishes that all life is genealogically related and that genealogical relationships form nested hierarchies entailed by the branching structure of evolutionary trees. But genealogical nesting and architectural incorporation are not the same claim, and only the second is what T&I specifically predicts.

Common descent is entirely consistent with evolutionary trajectories that dismantle earlier functional architectures and replace them with new ones. Path-dependence explains tinkering—the recurrent laryngeal nerve's detour, the inside-out retina, the repurposed jaw bones. It does not entail that each major transition must preserve the entire prior architecture as an autonomous functional subsystem with its own genome and replication machinery. Mitochondria retain near-complete metabolic independence and their own genomes, participating in regulatory relationships with the nuclear genome that could not exist at the prokaryotic level. This is not required by common descent plus path-dependence alone.

Maynard Smith and Szathm�ry's major-transitions research already describes these events in language mapping directly onto T&I: new levels of individuality with contingent irreversibility and mutual dependence. T&I is not inventing hierarchy—it is generalising a structural feature already visible in mainstream major-transitions literature. Similarly, the asymmetry T&I identifies at the level of holarchic transitions maps directly onto Dollo's Law—the accepted structural principle that complex structures, once lost, are not regained in identical form. T&I's irreversibility claim is of the same logical type as Dollo's Law. If the latter is scientifically legitimate, the former requires engagement rather than dismissal.

Human Developmental Stages

The developmental sequence documented independently by Piaget, Kohlberg, Loevinger, and Cook-Greuter exhibits the structural properties T&I specifically predicts: each stage incorporates rather than replaces the cognitive resources of the previous stage, the sequence is invariant in ordering across cultures, and earlier stages remain accessible and functional within later ones. This research was established entirely independently of Wilber's framework.

The mechanisms that implement this sequence—neural maturation, learning constraints, social scaffolding—explain how the sequence proceeds. They do not mandate the specific asymmetric inclusion structure T&I requires: that formal operational thinkers retain full functional access to concrete operational capacities rather than superseding them. Many cognitive architectures do not require this. The sharpest falsifiable version of the developmental prediction is a double criterion: any genuine post-formal stage must perform at least as well as formal-operational subjects on standard logical reasoning tasks, confirming inclusion of formal operations, while simultaneously demonstrating higher-order integration across perspectives that formal-operational subjects do not exhibit, confirming transcendence. This is testable with existing instruments and has not been disconfirmed in Cook-Greuter's independently established data.

The Consilience Argument

Taken individually, each domain observation faces partial alternative explanations. The more important question is whether a single structural principle—grounded in a necessity argument and formally instantiated in Cantor's transfinite hierarchy—generating observations consistent with its predictions across three mechanistically disconnected empirical domains constitutes evidence for that principle's reality beyond what individual domain explanations provide.

The domains share ontology but not mechanism. Mutation, selection, and endosymbiosis operate on one causal register; neural maturation, experience-dependent plasticity, and social scaffolding on another entirely. No direct causal chain links the retention of mitochondrial genomes to the invariance of Piagetian stage sequences. Whewell's consilience criterion requires mechanistic independence, not ontological separation. On that criterion the three domains satisfy it. What is striking is not merely that the three structural properties appear across biology, cognitive development, and cultural cognition, but that each domain independently conforms to what appears to be the only logically coherent structure of hierarchical expansion.

This is not a claim that T&I has been proven. It is a claim that it is a structural hypothesis with specified empirical commitments, identified failure conditions, and a necessity argument grounded in an independently established mathematical framework—deserving assessment on its own terms by the same logical structure as Darwin's argument for common descent, which was itself a consilience argument across independent domains rather than a formal statistical test.

Conclusion

Frank's charge that Eros lacks the features required of a scientific explanation is largely correct as applied to Eros as typically invoked. Where I part company is in the inference that this failing extends to T&I and the depth/span polarity when stated with the precision they deserve.

T&I is not a rival to natural selection. It characterises the invariant structural pattern that any adequate mechanistic account of hierarchical transitions must reproduce. The necessity argument shows that this pattern—inclusion, qualitative discontinuity, open-ended succession—is not one possible form hierarchical expansion might take, but the only logically coherent form it can take. The Cantor hierarchy is the formal mathematical instance of that structure. The biological and cognitive evidence is its empirical expression.

If that argument is wrong, the path to showing it is now clear: demonstrate how a genuine hierarchical transition could occur while violating inclusion, discontinuity, or open-endedness without dissolving the concept of a level-transition itself; or show that the falsification conditions I have stated are already met by existing evidence. I look forward to that more specific engagement.

[Frank Visser's response follows below.]