Taxonomy as Chiseling
分类即凿构：Self-as-an-End 的分类学理论

Author's Note

This paper is a methodology paper of the SAE theory framework, applying the framework's chisel, construct, and remainder concepts to a meta-analysis of taxonomic methodology itself. The complete theoretical foundation of the framework is found in three foundational papers: Paper 1 "Systems, Emergence, and the Conditions of Personhood" (Zenodo, DOI: 10.5281/zenodo.18528813), Paper 2 "Internal Colonization and the Reconstruction of Subjecthood" (Zenodo, DOI: 10.5281/zenodo.18666645), Paper 3 "The Complete Self-as-an-End Framework" (Zenodo, DOI: 10.5281/zenodo.18727327). The main case study taxonomy HC-16 is found in "HC-16: A Sixteen-Type Taxonomy of Subjective Vulnerability Profiles" (Zenodo, DOI: 10.5281/zenodo.18907264).

This paper is a philosophical analysis of taxonomic methodology, not an empirical study of any classification system. Case materials are used to illustrate structural arguments, not to establish empirical generalizations.

Chapter 1. The Problem: Why Taxonomy Needs a Meta-Theory

Core thesis: Every taxonomy chisels. Taxonomy is everywhere and rarely reflected upon. The need for a meta-theory of taxonomy is not merely philosophical but practically urgent.

1.1 Taxonomy Is Everywhere and Rarely Examined

Classification is one of the most basic operations of human cognition. Thinking itself is classification — distinguishing this from that, grouping the similar, imposing order upon diversity. The history of knowledge is, to a large extent, a history of classification acts: Aristotle's species taxonomy, Linnaeus's biological classification, Mendeleev's periodic table, Kant's Table of Categories, and in the domain of personality alone, the Big Five's five factors, the MBTI's sixteen types, attachment theory's four styles, and the recent HC-16's sixteen vulnerability profiles.

Yet despite this ubiquity, taxonomy rarely reflects on itself. Classification systems are created, disputed, applied, revised — but the cross-domain question "what must a good classification satisfy structurally" is rarely systematically foregrounded as an independent subject. Practitioners of specific classifications debate whether their dimensions are correct, whether their types are valid, whether their boundaries are appropriate. But the structural conditions a classification must satisfy — those are rarely clearly articulated as general inquiry.

1.2 Three Meta-Problems Revealed by Practice

The need for a meta-theory of taxonomy is not merely philosophical; it is practically urgent. Any multidimensional classification faces at least three meta-problems: the definition problem, the completeness-remainder tension, and the attribution problem. HC-16 (a sixteen-type taxonomy of subjective vulnerability profiles) recently encountered all three in its construction, bringing these already-ubiquitous problems to the foreground.

The definition problem. HC-16 originally defined "high" and "low" sensitivity as purely relative comparisons of four internal dimensions. Three independent reviewers immediately found a logical paradox: under a purely relative definition, all-high and all-low configurations are mathematically impossible. The solution requires introducing an axis baseline (relative to the general population's response, plus internal configuration). This is not HC-16's problem; it is taxonomy's problem. Any system classifying along multiple dimensions faces this: high and low relative to what?

The completeness-remainder tension. HC-16 claims framework-internal completeness (all 2⁴ = 16 combinations are generated), while simultaneously acknowledging a remainder (personality dimensions beyond the four pain dimensions). Reviewers marked this as contradictory. The solution requires distinguishing two levels: completeness within selected dimensions vs. the choice of dimensions itself. This is not HC-16's problem; it is taxonomy's problem. Any taxonomy claiming completeness while the world exceeds its dimensions faces this tension.

The attribution problem. HC-16 attributes the high frequency of some types to institutional calibration (competitive institutions systematically shaping people toward specific configurations). Reviewers asked: how to distinguish institutional calibration from evolutionary baseline? Without cross-cultural data from radically different institutional environments, this is non-falsifiable. This is not HC-16's problem; it is taxonomy's problem. Any taxonomy that explains its type distribution through environmental causes faces the challenge of separating environment from baseline.

These three problems — definition, completeness-remainder, attribution — are structural features of taxonomy itself. A meta-theory should identify them in advance and provide a handling framework.

1.3 Scope and Limits of This Paper

This paper addresses primarily the classification of complex, developable objects — especially personality, mental illness, organizations, and cultural configurations. For such objects, the full condition set developed (including inter-type dynamics, dynamism, and propagability) applies. For highly stable natural kinds or purely formal objects, some conditions (especially conditions four through six of Chapter 3) do not automatically apply; the core thesis (taxonomy as chiseling) and the Remainder Principle apply universally, but the colonization risk driving much of this paper's argument is specific to the classification of entities that may be harmed by misclassification.

This paper does not attempt to give a complete philosophy of classification. It advances one specific thesis — taxonomy as chiseling — and derives from it a set of structural conditions and methodological principles. The paper's own remainder — what it does not cover — includes the formal-mathematical theory of classification, the information-theoretic approach to classification, and the sociology of how classification becomes institutionalized. These are acknowledged as important but beyond the paper's scope.

Chapter 2. Taxonomy as Chiseling and Constructing

2.1 The Core Thesis

The core operation of the Self-as-an-End framework is chiseling and constructing. Chisel is the negating operation — exclusion, cutting, distinction, selecting what to attend to by deciding what to ignore. Construct is the affirmative operation — building, connecting, synthesizing, constructing structure from what the chisel exposes.

Classification is one of the most direct manifestations of these two operations.

A terminological note: this paper's "classification" includes both discrete typology (assigning objects to distinct categories) and dimensional models (locating objects in a continuous dimensional space or configuration space). "Type" can refer to discrete categories or to stable regions in a configuration space. The Big Five is a dimensional model; MBTI is a discrete typology; HC-16 is at an intermediate position (discrete types defined in a four-dimensional configuration space). The following analysis applies to all three.

Every classification act begins with chiseling: selecting classification dimensions. Selecting "these dimensions" necessarily excludes "other dimensions." The Big Five chisels from factor analysis (statistical clusters emerging from personality-descriptive adjectives). MBTI chisels from Jungian typological intuition (Jung's clinical observation of important preference pairs). HC-16 chisels from four structural pain dimensions (sensitivity configurations produced by the cross of base-layer/emergent-layer and catalytic-pain/driven-pain). Each chisel illuminates a specific facet of the object while casting other facets into shadow.

Every classification act continues with constructing: the types generated by the chisel are given names, definitions, portraits, relations, and dynamics. Construct makes classification usable — it transforms a set of abstract dimension combinations into a cognitive tool that people can use to understand themselves, understand others, and predict behavior.

Chisel and construct are not independent. Chisel determines what construct can see; construct determines what chisel is for. A chisel without construct is a cut that doesn't know where it is going. A construct without awareness of its own chisel is a system that has forgotten its own origin — and this forgetting is precisely where the fundamental risk lies.

2.2 The Fundamental Risk: Forgetting the Chisel

The deepest danger of taxonomy is not misclassification. Misclassification can be corrected through better data, more precise tools, finer observation. The deepest danger is forgetting that one has chiseled — forgetting that a classification is a perspective from a particular angle, not the object itself.

When a classification forgets its own chisel, its construct begins claiming totality. "Personality just is these five dimensions." "You are one of these sixteen types." "Human disorders divide into these categories." This claim treats the construct as having no remainder — as if the selected dimensions were the only existing dimensions, as if the generated types were the only possible types, as if what the classification cannot see does not exist.

In the language of the Self-as-an-End framework, this forgetting is the starting point of colonization. The construct colonizes the object by claiming to exhaust it. The remainder — everything the chisel excluded — is denied rather than acknowledged. Classification transforms from a tool for understanding into a tool for control: if you do not fit, the problem is with you, not with the classification.

Every major classification in history has experienced this colonization to some degree. DSM diagnostic categories were originally designed as clinical convenience tools and later reified as discrete natural kinds. MBTI's sixteen types were originally offered as self-exploration tools and later used as hiring criteria. The Big Five's five factors were originally extracted as statistical summaries and later treated as the basic structure of personality. In every case, the construct forgot the chisel that produced it, and the remainder was suppressed.

2.3 The Remainder Principle

From the core thesis, a proposition that may be called the Remainder Principle can be derived:

The remainder is systematic, not random. It is not "a few people who don't fit well." It is a class of differences that is systematically invisible from this angle — a kind of difference that cannot be seen from this perspective. The remainder is not the residue left when the classification does badly enough; it is the invisible domain systematically produced by the chosen chisel.

MBTI's remainder is not "a few people whose type is hard to determine"; it is all people whose core personality differences do not fall on the E/I, S/N, T/F, J/P dimensions. HC-16's remainder is not "a few people whose vulnerability profile is unclear"; it is all people whose core differences fall on cognitive style, temperament, or value orientation rather than on structural pain sensitivity.

The Remainder Principle has a practical implication: the remainders of different classifications may be mutually complementary. What classification A cannot see, classification B may illuminate, and vice versa. This is why taxonomic pluralism — the coexistence of multiple different-chisel classifications — is not a sign of failure but a structural necessity. No single chisel can see everything. Pursuing a single, totalizing classification is itself a colonizing impulse.

2.4 Case Study: The Chisel of MBTI and Its Structural Remainder

MBTI provides an illuminating case study — a taxonomy whose extraordinary propagation success has masked a structural problem in its dimensional selection.

MBTI's four preference pairs — Extraversion/Introversion, Sensing/Intuition, Thinking/Feeling, Judging/Perceiving — derive from Jungian typological intuition, developed by Katharine Briggs and Isabel Myers. This chisel cuts along cognitive and behavioral preference lines: how you direct energy (E/I), how you acquire information (S/N), how you make decisions (T/F), how you orient to the external world (J/P).

This chisel is extraordinarily effective for propagation. The dimensions cut along facets of self-experience that most people can readily recognize — most people can say whether they prefer introversion or extraversion, thinking or feeling in decisions. This immediate recognizability is the engine of MBTI's viral propagation.

But this chisel has a structural problem that becomes visible when you try to map its dimensions onto any structurally derived framework. Consider the S/N pair. Sensing (S) refers to direct sensory data in the present moment — a single-layer capacity. Intuition (N), per MBTI's description, refers to pattern recognition beyond immediate sensory data, possibilities and meaning — a cross-layer capacity, not within any single layer. MBTI pairs a single-layer capacity with a cross-layer capacity and treats them as two poles of a single dimension.

This structural misfit provides a possible structural explanation for an empirical observation: S/N is the dimension with the lowest test-retest reliability in MBTI. People's S/N scores change most frequently on repeated administration. The standard explanation is that people are "unclear" about their S/N preference. If S and N are indeed not the two poles of a single structural dimension, then the relevant dimension's instability need not be entirely attributed to respondents being "unclear about themselves" — it may partly come from the dimension itself.

This analysis is not an attack on MBTI. It is a demonstration: when classification dimensions derive from clinical intuition rather than structural analysis, the chisel may cut along lines that do not correspond to structural joints, producing internally inconsistent dimensions. MBTI's chisel illuminates real and important facets of personality (its propagation success is proof), but it cuts along structurally oblique lines — across structural layers rather than along them.

Chapter 3. The Structural Conditions of Good Taxonomy

The core thesis — taxonomy as chiseling — generates a two-level structure of conditions that good taxonomies must satisfy.

3.1 The Ontological Condition: Acknowledging the Remainder

The most fundamental condition is not technical but ontological: a good taxonomy knows it has a remainder.

This condition is prior to all others because it determines the taxonomy's relationship to its own status. A taxonomy that acknowledges its remainder treats itself as a tool — powerful, illuminating, but partial. A taxonomy that denies its remainder treats itself as a mirror of reality — complete, definitive, total. The former invites complementation and revision; the latter resists them.

The ontological condition is not merely a disclaimer to be appended ("of course no taxonomy is perfect"). It must be structurally built into the taxonomy itself. HC-16, for instance, devotes an entire section (2.5) to explicitly marking what its chisel illuminates and what it occludes, and a full chapter (Chapter 6) to analyzing its remainder and the colonization risks that follow from forgetting it. The Big Five, by contrast, has no structural acknowledgment of remainder — its five factors are presented as the fundamental dimensions of personality, with the implicit claim that personality variation not captured by these five factors is noise rather than signal.

3.2 The Technical Conditions

Below the ontological condition, six technical conditions define what makes a taxonomy well-built. A note on layering: conditions one through three (structural grounding, framework-internal completeness, discriminative power) are core conditions shared by most taxonomies across domains. Conditions four through six (inter-type dynamics, dynamism, propagability) apply primarily to taxonomies of complex, developable objects intended to function as public cognitive tools.

Condition one: structural grounding of dimensions. The classification dimensions should have identifiable sources — whether from structural derivation (Kant's categories from the logical functions of judgment; HC-16's four pain dimensions from the 2×2 matrix of layers and transmission directions), from empirical induction (the Big Five's factors from adjective clustering), or from clinical observation (attachment theory's dimensions from the Strange Situation). The key requirement is not that dimensions must be a priori, but that they must be traceable — one must be able to answer "why these dimensions and not others?" If the answer is "because these are the ones we happened to measure," the grounding is weak. If the answer is "because these follow from a structural analysis of the domain" or "because these are the dimensions that consistently emerge across multiple independent datasets," the grounding is strong.

Condition two: framework-internal completeness. Within the chosen dimensions, the combinatorial space should be exhausted. If a taxonomy has four binary dimensions, all 2⁴ = 16 combinations should be acknowledged, not just the ones that seem empirically common. Completeness is always framework-internal — "within these dimensions, nothing is left out" — and must never be confused with domain-completeness — "everything in the domain is captured." The former is a legitimate structural claim; the latter is a colonizing overclaim.

Condition three: discriminative power. The types generated by the taxonomy should differ from each other in ways that matter for some purpose. "Mattering" is purpose-relative: the Big Five's dimensions matter for predicting job performance; the MBTI's types matter for self-recognition and interpersonal understanding; HC-16's types matter for diagnosing vulnerability and prescribing repair. A taxonomy whose types do not differ in any consequential way — whose classification produces distinctions without differences — fails this condition regardless of its other virtues.

Condition four: inter-type dynamics. For taxonomies that aim to explain multi-agent interaction or relational consequences, inter-type dynamics is a decisive condition. A good taxonomy of this kind does not stop at listing types. It provides an account of what happens when types interact — complementarity, conflict, resonance, blind-spot stacking, or other patterns. Without dynamics, a taxonomy is a collection of specimens pinned to a board. With dynamics, it becomes an ecology.

Condition five: dynamism. Types should not be lifelong sentences. A good taxonomy of persons, organizations, or cultures builds in mechanisms for type change — what moves a classified entity from one type to another, under what conditions, and through what process. Taxonomies of natural kinds may legitimately be static — gold does not become silver. But taxonomies of any entity that develops over time must build in dynamism or risk colonizing their objects into fixed identities.

Condition six: propagability. Propagability is not a condition of scientific truth but a condition of cognitive-tool function — it applies when a taxonomy aims to be used by the public, not merely by specialists. A taxonomy that cannot be remembered, communicated, and discussed cannot fulfill its purpose. Propagability is a function of naming (can types be named in memorable, pronounceable ways?), visual representability (can the taxonomy be displayed as a table, map, or diagram?), and self-identification ease (can a person find their own type without expert assistance?).

A critical caveat: propagability stands in structural tension with colonization risk. The more propagable a taxonomy, the more people it reaches — but also the higher the probability that types are reified into identity labels rather than used as diagnostic tools. Good taxonomy does not maximize propagability unconditionally. It pursues propagability subject to the constraint that structural precision is not sacrificed — that the simplification required for memorability and communicability does not cross the threshold into distortion.

3.3 The Relationship Between Levels

The ontological condition and the six technical conditions are not independent. The ontological condition — acknowledging the remainder — constrains how the technical conditions are interpreted.

A taxonomy that satisfies all six technical conditions but denies its remainder is technically excellent and ontologically colonized. It classifies well within its chosen dimensions but claims (implicitly or explicitly) that its dimensions exhaust the domain. Many successful taxonomies fall into this category — they are powerful tools that have forgotten they are tools.

A taxonomy that acknowledges its remainder but fails the technical conditions is ontologically honest and technically useless. It knows it is partial but does not classify well even within its partial scope.

The goal is both: ontological honesty (knowing one has chiseled) and technical rigor (chiseling well). These are not in tension; they are complementary. Knowing your remainder makes your construct more honest; building your construct well makes your remainder more precisely identifiable.

Chapter 4. The A Priori and A Posteriori in Taxonomy

4.1 Two Routes, One Destination

The history of taxonomy shows two major routes to classification.

The a priori route derives classification dimensions from structural principles before looking at empirical data. Kant's Table of Categories is the paradigmatic case: twelve categories derived from four logical functions of judgment, with completeness guaranteed by the derivation. HC-16 follows this route: four pain dimensions derived from a 2×2 structural matrix, with sixteen types generated combinatorially.

The a posteriori route extracts classification dimensions from empirical data. The Big Five is the paradigmatic case: five factors extracted from factor analysis of personality-descriptive adjectives, with the number and identity of factors determined by the data. Linnaeus's biological taxonomy similarly begins from observed morphological similarities and differences.

The standard framing treats these as competing approaches — one must choose between theory-driven and data-driven classification. This paper argues that the framing is wrong. For taxonomies of complex, developable objects — the primary scope of this paper — good taxonomy requires both, in a specific structural relationship.

4.2 How the A Priori and A Posteriori Constrain Each Other

The a priori constrains the a posteriori by providing a structural skeleton — a set of expectations about what dimensions should exist and what combinations should be possible. When HC-16's a priori derivation generates sixteen types, it tells the empirical researcher: "look for these sixteen configurations; if you find fewer, ask why the missing ones are absent rather than assuming they don't exist." The a priori provides search direction.

The a posteriori constrains the a priori by providing reality checks — evidence about whether the structurally derived types actually exist in the world, how they are distributed, and whether the predicted dynamics hold. If HC-16's sixteen types are administered to large populations and certain types are consistently unfindable, this constrains the a priori: perhaps the relevant dimension is not empirically independent, or perhaps the high/low threshold is drawn in the wrong place. The a posteriori provides empirical correction.

The relationship is cyclical, not one-directional. The a priori generates expectations; the a posteriori tests them; anomalies feed back to refine the a priori; the refined a priori generates new expectations; and so on. Good taxonomy lives in this cycle. Bad taxonomy breaks it — either by refusing empirical correction (pure a priori dogmatism) or by refusing structural interrogation (pure a posteriori empiricism).

4.3 The Pathologies of Purity

Pure a priori taxonomy derives its types from structural principles and never asks whether they correspond to anything in experience. The risk is empty formalism — a taxonomy that is internally elegant but externally vacant. Some of Kant's categories have been criticized on precisely this ground: the category of "community" (reciprocal causation), while derivable from the logical function of disjunctive judgment, has been questioned as to whether it picks out a genuine structure of experience or merely fills a slot in a table.

Pure a posteriori taxonomy extracts its types from data and never asks why these types and not others. The risk is structural arbitrariness — a taxonomy whose dimensions are contingent on the particular dataset, language, or population from which they were extracted. The Big Five's five factors are stable across many datasets, but the question "why five?" has no structural answer. In some languages and cultures, factor analysis yields six factors; in others, three. The Big Five's defenders argue that five is the most robust solution; its critics point out that "most robust" is not the same as "structurally necessary."

The co-presence of both routes avoids both pathologies. A priori derivation ensures that the dimensions have structural reasons for existing; a posteriori testing ensures that the structurally derived types have empirical traction. Neither alone is sufficient.

4.4 HC-16 as a Case Study of Co-Presence

HC-16 illustrates both the promise and the unfinished business of co-presence. The a priori side: four pain dimensions derived from a 2×2 matrix, sixteen types generated combinatorially, framework-internal completeness guaranteed. This side is fully developed. The a posteriori side: currently consists of theoretical expectations about type frequency, portraits based on structural extrapolation, and predictions with falsification conditions. This side is explicitly unfinished — the paper's open questions (empirical validation strategy, remainder identification, MBTI complementarity test) are precisely the a posteriori agenda that remains to be executed.

HC-16 is thus at a specific point in the a priori / a posteriori cycle: the a priori derivation is complete, and the a posteriori calibration has been designed but not yet performed. This is an honest position — far better than either claiming empirical validation that has not been done, or dismissing the need for it.

Chapter 5. The Colonization and Cultivation of Taxonomy

5.1 How Taxonomies Colonize

Every taxonomy, once established, faces the risk of colonization — the process by which a cognitive tool degrades into a control instrument. Taxonomic colonization takes three characteristic forms.

Type reification: types, originally understood as regions in a configuration space, are treated as fixed natural kinds. "You are an INFP" ceases to mean "your current configuration falls in this region" and begins to mean "this is your essence." The type absorbs the person; the person becomes the type.

Remainder denial: the taxonomy's blind spots are treated as non-existent rather than as consequences of a particular chisel. If the taxonomy cannot see it, it does not exist. People who do not fit are pathologized ("your results are inconsistent") rather than recognized as the taxonomy's remainder.

Institutional capture: the taxonomy is adopted by institutions (corporations, schools, governments) as a sorting mechanism. Types are used to determine eligibility, allocate resources, or restrict possibilities. The MBTI used as a hiring filter. Psychiatric diagnoses used as gatekeeping criteria. Personality types used to stream students into educational tracks. In every case, the taxonomy becomes a mechanism for controlling people rather than understanding them.

5.2 How Taxonomies Cultivate

Taxonomies can also serve a cultivating function — helping subjects understand their own structural conditions and thereby enabling more informed self-navigation.

The minimal cultivation function is self-recognition: "this is my configuration." The MBTI's extraordinary popularity rests largely on this function — the "that's so me" moment is a micro-instance of cultivation. The subject perceives their own structure and experiences the recognition as liberating rather than constraining.

The extended cultivation function is self-diagnosis: "this is where I am most vulnerable." HC-16 is designed to provide this function — each type portrait includes not just "what you are like" but "where you are most likely to fall" and "what works for you." The taxonomy becomes a diagnostic tool that points toward repair.

The deep cultivation function is structural self-awareness: "I know that I am this configuration, I know that this configuration is not my essence, and I know that knowing this changes how the configuration operates." This is the metacognitive level at which taxonomy most powerfully serves the subject — not by giving them a label but by giving them a structural map that includes the instruction "this map is not the territory."

5.3 The Structural Design Against Colonization

Taxonomies can be designed to resist colonization. Several structural features serve this function:

Built-in dynamism: if the taxonomy demonstrates that types move over time, type reification becomes harder to sustain. HC-16's Chapter 5 (institutional, relational, and self-calibration) is a structural argument against treating types as essences.

Explicit remainder acknowledgment: if the taxonomy openly states what it cannot see, remainder denial becomes harder to sustain. HC-16's Section 2.5 and Chapter 6 serve this function.

Prohibition on institutional sorting: if the taxonomy explicitly states that it must not be used for hiring, promotion, educational tracking, or partner selection, institutional capture is at least formally resisted. HC-16 includes this prohibition. Whether formal prohibition is sufficient to prevent institutional capture in practice is an open question — the MBTI was never designed for hiring, but it is widely used for hiring nonetheless.

Prescription function: if the taxonomy points toward repair rather than fixation — if "you are this type" leads to "here is what to attend to" rather than "here is what you are stuck with" — the taxonomy orients toward action rather than stasis. The presence of a prescription function transforms the taxonomy from a mirror (reflecting what is) into a compass (pointing toward what could be).

Chapter 6. Non-Trivial Predictions

The following predictions rely on several working indicators: cross-cultural stability is measured by factor replication rates or type-distribution robustness across cultures; self-report accuracy is measured by test-retest consistency and self-observer convergence; type-switching frequency is measured by the rate of self-assigned type change across repeated long-term administrations; cultivation/colonization orientation is measured by changes in self-insight value endorsement versus constraining-effect endorsement.

6.1 A Priori / A Posteriori Convergence Predicts Cross-Cultural Stability

6.2 Built-In Dynamics Increases Self-Report Accuracy

6.3 Propagability and Structural Grounding Inversely Predict Type-Switching Frequency

6.4 Institutional Adoption Degrades Cultivation Function

Conclusion

This paper proposed a meta-theory of taxonomy grounded in the Self-as-an-End framework's chisel, construct, and remainder concepts.

The core thesis is that classification is a chiseling act: every classification selects dimensions (chisels), and in these dimensions builds types, names, and relations (constructs). The fundamental risk is not misclassification but forgetting that one has chiseled — claiming to have no remainder. This forgetting is the starting point of taxonomic colonization (Chapter 2).

From this, a two-level condition structure is derived. Ontological level: a good taxonomy acknowledges its remainder. Technical level: structural grounding, framework-internal completeness, discriminative power, inter-type dynamics, dynamism, and propagability. The ontological condition constrains how all technical conditions are interpreted (Chapter 3).

The paper argued that good taxonomy requires the co-presence of a priori and a posteriori — the former provides structural skeleton, the latter provides empirical calibration, and both constrain each other in a continuous cycle. Pure a priori taxonomy risks empty formalism; pure a posteriori taxonomy risks structural arbitrariness. HC-16 is analyzed as a case study at a specific position in this cycle: a priori derivation complete, a posteriori calibration designed but not yet executed (Chapter 4).

Taxonomic colonization and cultivation functions were analyzed, and anti-colonization structural design features identified (Chapter 5). Four non-trivial predictions with falsification conditions were derived (Chapter 6).

This paper's contribution to the Self-as-an-End framework is applying chisel, construct, and remainder to a new domain — taxonomic methodology itself. For analyzing subjecthood and its colonization, the framework concepts prove applicable to any classification enterprise: classifications, like subjects, can be colonized when their remainder is denied.