Formalization¶
Core Idea¶
Formalization is the deliberate process of rendering informal, tacit, or implicit practice into explicit, codified, rule-governed form — notation, axioms, statutes, schemas, or standards — such that what was previously carried by intuition, habit, or convention becomes statable, checkable, and transmissible. The defining commitment is a move up the explicitness gradient: replacing "we just know how" with an articulated system whose elements and inference rules are laid out and can be operated on mechanically or audited against. [1] The intellectual high-water mark of this move is the early-twentieth-century formalist program in mathematics, where Hilbert (1899/1902) re-grounded Euclidean geometry on a set of explicit, gap-free axioms precisely to expose every assumption that classical practice had carried silently. [2]
What unifies formalization across domains is not the content being codified but the direction of the move and its characteristic payoff structure: it surfaces hidden assumptions, enables mechanical checking and automation, and makes a practice portable beyond the people who originally held it — at the cost of freezing what was once fluid and shedding tacit nuance that resisted statement. The act is intentional and human-driven: someone decides that a practice has become stable, important, or contested enough to be worth writing down, and undertakes the labor of articulating it. [3] Crucially, formalization concerns knowledge and practice, not matter or energy; it is a move within the space of how-to and what-is-the-rule, which is why it appears wherever humans accumulate competence they wish to transmit, audit, or hand off to a machine.
How would you explain it like I'm…
Writing the rules down
Making hidden rules explicit
Codifying tacit practice into rules
Structural Signature¶
Formalization encodes a structural pattern: tacit/implicit practice → articulation of elements and rules → explicit checkable system → mechanical operation or audit. It separates two regimes (a practice carried by intuition and a practice stated as a system) and names the deliberate transformation that crosses between them, fixing terms, enumerating cases, and laying out inference or procedure so that conformance can be judged without re-deriving the practice from scratch. [3]
Recurring features:
- Move up the explicitness gradient from tacit to stated
- Deliberate codification of practice into explicit rules
- Replacing "we just know how" with an articulated system
- Making assumptions statable, checkable, and transmissible
- Fixing terms and inference so conformance can be audited
- Trading flexibility for reliability and mechanical operability
- Rendering know-how into an artifact independent of its holder
The structural insight is robust: a logician translating an intuitive argument into a symbolic calculus, a legislature converting customary practice into statute, an engineering team turning a release ritual into a written runbook, and a knowledge engineer encoding an expert's diagnostic intuition into a ruleset all execute the same move. [3] Each one trades the fluid, person-bound, context-sensitive version of a practice for a fixed, inspectable, transmissible one — and each inherits the same recurring liability, that the explicit system can fail to capture the judgment the tacit version supplied.
What It Is Not¶
Formalization is not a claim that the formal system is truer or better than the practice it codifies. The codified version is a model — a deliberate compression that fixes terms and rules — and like any model it can be faithful, lossy, or actively misleading relative to the messy practice it abstracts. [1] Naming the prime does not assert that what gets written down is right; it only names the act of writing it down in a checkable form. A statute can codify a bad custom; an ontology can entrench an expert's blind spots.
Nor does formalization claim that more explicitness is always better. The recurring lesson across domains is the opposite: over-formalization ossifies, freezing practices that needed to keep evolving and substituting brittle rule-following for the situated judgment that made the original practice work. The prime is direction-neutral about how far to go; it names the move, not an injunction to maximize it. There is a real and contested optimum, and pushing past it is a characteristic failure mode, not a success.
Formalization is also not mere documentation or description. Writing down what happened, or describing a practice in prose, does not by itself formalize it. Formalization specifically produces a rule-governed artifact — one with stated elements and inference or procedure such that conformance can be mechanically judged or audited. A field journal describing how an expert diagnoses faults is documentation; a decision tree or ruleset that a machine can execute to reach the same diagnosis is formalization. The differentiator is operability: the explicit system must be something you can run against a case, not merely read.
Finally, the prime makes no claim that formalization is a one-time terminal event. Codified systems are revised, amended, deprecated, and re-formalized; a constitution is amended, a spec is versioned, an axiom set is found inconsistent and repaired. Formalization names a recurring move within a living system, not the construction of a permanent monument.
Broad Use¶
Logic: Translating an intuitive argument into a symbolic system where validity is decidable by form alone — the paradigm case, in which Frege's (1879) Begriffsschrift introduced a notation explicit enough to make inference itself an object of inspection rather than an exercise of intuition. [4]
Mathematics: Axiomatizing a body of informal results so that theorems follow by stated rules — Euclid's geometry, Peano arithmetic, Zermelo–Fraenkel set theory — each replacing a working mathematician's sense of "obviously true" with explicit premises whose consequences can be derived and checked.
Law: Codifying customary practice into written statute, converting "what is usually done" into binding rule — the move from common-law accretion to codes such as the Napoleonic Code, which sought to state the law explicitly enough that a citizen could read their obligations rather than infer them from precedent. [5]
Organizational theory: Turning informal routines into documented procedures, policies, and org charts — Weber's (1922) analysis of bureaucratic rationalization treats this codification of office, rule, and jurisdiction as the engine that lets large organizations operate impersonally and at scale. [6]
Knowledge engineering (non-obvious): Encoding a domain expert's tacit know-how into an ontology, ruleset, or schema a machine can apply — the central project of expert systems and, more recently, knowledge graphs, where the bottleneck has long been precisely that experts cannot fully state what they know. [7]
Standards: Crystallizing de facto practice into a published specification — file formats, network protocols, programming-language grammars — converting "this is how everyone happens to do it" into a single authoritative reference against which implementations can be conformance-tested.
Clarity¶
Naming formalization lets practitioners see the act of making explicit as a distinct move with its own costs and gains, rather than as a neutral or automatic by-product of maturity. [3] It separates two questions that are easily conflated: "do we have a working practice?" and "do we have a stated system for the practice?" An organization can be highly competent yet entirely unformalized (everything lives in people's heads), and conversely it can be heavily formalized yet incompetent (the manual is thick and the work is bad). Pulling these apart redirects attention from "are we good at this?" to "what is gained and lost by writing this down, and how far should we go?"
The concept also clarifies a characteristic surprise: the act of formalizing routinely changes the practice it was supposed to merely record. The logician formalizing an argument discovers hidden premises; the legislature codifying custom is forced to decide explicitly what the custom had left ambiguous, resolving by fiat questions that the informal version had productively left open. Polanyi's (1966) thesis that "we can know more than we can tell" gives the sharpest statement of why this happens: a substantial residue of any skilled practice is tacit, and the demand to state it forces choices that the tacit version never had to make. [1] Clarity here is the recognition that formalization is not a transcription but a transformation.
Manages Complexity¶
Formalization compresses scattered, person-bound know-how into a shared, inspectable artifact that can be reused, taught, audited, and mechanically applied without re-deriving it each time. It bounds ambiguity by fixing terms and rules, trading flexibility for reliability and transmissibility — the same trade that lets a compiler reject malformed programs, a court rule on conformance to statute, and an auditor check a process against documented procedure. [3] Once a practice is formalized, the cognitive load of executing it drops: the rule can be followed without reconstructing the reasoning that produced it, and disagreements can be settled by pointing at the text rather than relitigating intuitions.
This is also where formalization's complexity-management payoff turns against itself. A formal system that has fixed the wrong terms, or fixed terms that the world later outgrows, becomes a source of complexity rather than a sink for it: the organization now maintains both the messy reality and the formal artifact that no longer matches it, and must continually translate between them. Managing complexity through formalization therefore carries an ongoing maintenance cost — the explicit system must be kept aligned with the evolving practice, or it silently accumulates the gap between map and territory that every codified system tends to grow.
Abstract Reasoning¶
Recognizing formalization supports reasoning about the explicit/tacit trade-off — what is lost when intuition is codified — and about the gap between the formal model and the messy practice it abstracts. [1] It enables a distinctive counterfactual move: "what would we be forced to decide if we had to write this down?" Asking this of an informal practice surfaces its hidden assumptions and unresolved cases before any artifact is built, which is why drafting a spec is so often diagnostic even when the spec is never finished.
The prime also licenses reasoning about when to formalize. The transfer-bearing heuristic is that formalization pays off for practices that are stable, high-stakes, and frequently transmitted — where the cost of articulating once is repaid by many checkable, hand-off-able uses — and backfires for practices that are fast-changing or judgment-heavy, where the artifact ossifies faster than it is amortized. This lets a reasoner predict, across unfamiliar domains, where codification will help and where it will hurt, by asking about the stability and transmission-frequency of the practice rather than about the domain itself.
Knowledge Transfer¶
The logician's experience that formalizing an argument exposes hidden premises transfers directly to the lawyer codifying custom (latent exceptions surface and must be adjudicated) and the engineer writing a spec (edge cases become explicit and demand a decision). [1] The structural lesson — articulation forces resolution of what was left implicit — is the same in each, even though the substrate is an argument, a custom, or a protocol. A practitioner who has lived through one such formalization can anticipate the others: that the act will take longer than expected, that it will unearth disagreements no one knew existed, and that the resulting artifact will need maintenance.
The organizational lesson that over-formalized procedure ossifies transfers to the knowledge engineer warned that a rigid ontology can fail to capture expert judgment, and to the standards body warned that an over-specified protocol can foreclose the innovation that the informal practice would have permitted. The shared transfer is the cost side of the same trade: every gain in checkability and transmissibility is bought with a loss in flexibility and tacit responsiveness, and a practitioner who carries this lesson from one domain arrives in the next already asking the right question — not "should we formalize?" as a yes/no, but "how far, and what will we lose?"
Examples¶
Formal/abstract¶
Axiomatizing geometry: For two millennia, geometers reasoned from Euclid's Elements treating its propositions as self-evident, while quietly relying on intuitions the text never stated — that a line drawn between a point inside and a point outside a circle must cross the circle, for instance. Hilbert's Foundations of Geometry (1899) re-derived the whole edifice from an explicit, complete set of axioms governing incidence, order, congruence, parallels, and continuity, exposing exactly these silent assumptions and making it possible to ask precisely which theorems depend on which axioms. The informal practice (drawing figures and reasoning about them) became a formal system (deriving consequences from stated premises by stated rules). Mapped back: This is the prime in its purest form — a move up the explicitness gradient that surfaced hidden premises, enabled mechanical checking of which results survive when an axiom is dropped, and made geometric knowledge transmissible as a system rather than a craft. The cost was equally characteristic: the formal version sheds the figural intuition that made geometry humanly tractable, which is why students still learn the informal practice first.
Formalizing an inference: A working scientist argues that since the treatment group improved and the control group did not, the treatment caused the improvement. Rendering this into an explicit inferential framework forces articulation of every premise the intuitive argument glossed: randomization, the absence of confounds, the meaning of "caused," the statistical model linking observation to conclusion. Premises that the informal argument carried silently must now be stated and either defended or discharged. Mapped back: The structure is identical to axiomatization — articulation of elements and rules transforms a tacit practice (judging that an argument is sound) into an explicit system (a stated chain of inference whose validity can be checked by form). And as always, the act is diagnostic: most of the value arrives not in the finished formalization but in the hidden premises it forces into the open.
Applied/industry¶
Codifying custom into statute: A community has long settled disputes by the judgment of elders, who weigh circumstances case by case. Writing this practice down as a statute forces the community to decide explicitly what was once handled by discretion: which factors count, in what order, with what weight, and what the fixed remedies are. The result gains consistency, transmissibility, and the ability to bind officials who never apprenticed under the elders — and loses the situated discretion that let the elders treat each case on its merits. Mapped back: This is formalization in the institutional substrate: tacit practice (elders' judgment) becomes an explicit, checkable system (statute) that can be applied by anyone and audited against. The recurring liability appears on schedule — once codified, the rule binds even cases the elders would have treated as exceptions, the classic ossification cost of pushing formalization past its useful point.
Writing the runbook: An engineering team has a release ritual that lives in the senior engineer's head: a sequence of checks, environment toggles, and "watch out for X on Tuesdays" lore. Turning it into a written runbook — and then into an automated pipeline — forces every implicit step and edge case to be stated, sequenced, and made conditional, so that a new engineer or a script can execute the release without the senior engineer present. The same demand for explicitness drives the encoding of an expert's diagnostic intuition into an expert system's ruleset, where the bottleneck is precisely that the expert cannot fully say what they know. Mapped back: Both cases run the prime in the computational/organizational substrate: person-bound tacit know-how is articulated into a rule-governed artifact that can be mechanically operated and handed off. And both show the boundary of the move — the parts of the practice that resist statement (the senior engineer's feel for when something is "off," the expert's gestalt judgment) are exactly what the formal version struggles to capture.
Structural Tensions¶
T1: Articulation reveals, but it also distorts. Formalization is prized because writing a practice down surfaces hidden assumptions and forces unresolved cases into the open. But the very act of stating what was tacit changes it: questions the informal practice productively left open must now be answered by fiat, and the answers chosen become binding even where the original judgment would have varied. The instrument that reveals the structure also deforms it, and there is no formalization that purely transcribes.
T2: The optimum is real but unmarked. Under-formalization leaves a practice trapped in people's heads, untransmissible and unauditable; over-formalization ossifies it, freezing what needed to evolve and substituting brittle rule-following for judgment. There is a genuine optimum between these, but nothing in the act of formalizing tells you where it is. Practitioners routinely overshoot, because each marginal codification looks locally like an improvement in rigor while the cumulative loss of flexibility is diffuse and shows up only later.
T3: Formalization both democratizes and entrenches power. Writing the rule down lets anyone read their obligations and hold officials to the stated text rather than to unaccountable discretion — a democratizing move. Yet the act of codification is itself an exercise of authority: whoever drafts the statute, the spec, or the ontology fixes the terms in which the practice will be conducted thereafter, and embeds their assumptions where they are hard to dislodge. The same artifact that constrains the powerful also enthrones the drafter.
T4: Checkability is bought with the loss of tacit nuance. The payoff of formalization is that conformance can be judged mechanically, without re-deriving the practice. But the practice's tacit residue — the expert's feel, the elder's sense of the case, the senior engineer's intuition for when something is off — is exactly what cannot be reduced to a checkable rule. The dimensions that make a practice good are often the dimensions that resist formalization, so the formal system tends to optimize the measurable and quietly abandon the rest.
T5: A formal system is a model that can drift from its subject. Once codified, the explicit artifact and the living practice are two things, and they diverge: the world changes, the practice adapts, and the formal system lags unless actively maintained. The organization then carries both the messy reality and the artifact that no longer matches it, paying an ongoing translation cost. Formalization promises to manage complexity, but an unmaintained formal system becomes a new source of it.
T6: Formalizing can foreclose the innovation the informal version permitted. An informal practice tolerates variation, local experiment, and quiet deviation — much of which is where improvement comes from. An over-specified standard or rigid procedure can lock in the practice as it was at the moment of codification, making the very deviations that would have improved it into violations. The reliability that formalization buys is, in part, the suppression of the exploration that an informal practice was running for free.
Structural–Framed Character¶
Formalization sits toward the structural side of the structural–framed spectrum, with some framing: it is the process of rendering informal, tacit, or implicit practice into explicit, codified, rule-governed form — notation, axioms, statutes, schemas — so that what was carried by intuition or habit becomes statable, checkable, and transmissible. The defining move is up the explicitness gradient: "we just know how" becomes an articulated system.
The core process is neutral and carries no evaluative weight, and applying it recognizes a real shift from tacit to explicit rather than importing a view — visible when a mathematician axiomatizes an informal argument or a logician formalizes an inference rule. What adds mild framing is that its canonical targets are human practices: turning custom into statute or routine into written procedure presupposes a community whose practice is being codified, so a partial institutional referent comes along. Recognition and neutrality read structural; the human-practice targets supply the framing.
Substrate Independence¶
Formalization is a highly substrate-independent prime — composite 4 / 5 on the substrate-independence scale. Its core — a move up the explicitness gradient that replaces tacit know-how with a statable, checkable, mechanically operable system — is substrate-agnostic and carries real design leverage. It spans the formal substrate of logic and mathematical axiomatization, the social-institutional codifying of custom into statute and procedure, and the computational act of writing a spec, with the recurring lesson that over-formalization can ossify. It holds breadth at 4 because it is largely absent from physical and biological substrates, concerning knowledge and practice rather than matter.
- Composite substrate independence — 4 / 5
- Domain breadth — 4 / 5
- Structural abstraction — 4 / 5
- Transfer evidence — 4 / 5
Relationships to Other Primes¶
Parents (2) — more general patterns this builds on
-
Formalization presupposes Representation
Formalization is the move up the explicitness gradient — replacing tacit know-how with codified notation, axioms, or statutes — and that move only operates inside a representational medium. To articulate a previously implicit rule one needs symbols, schemas, or formal language that map onto the practice being captured under a faithfulness convention. Representation supplies the structured mapping of target to medium that formalization then sharpens into mechanically operable, audit-checkable form, so formalization cannot get off the ground without a representational substrate already available.
-
Formalization is a decomposition of Transformation
Formalization is the particularization of transformation to the explicitness gradient: the input is tacit, conventional, or intuitive practice; the rule is articulation into notation, axioms, statutes, or schemas; the output is explicit, checkable, transmissible system. Where transformation names structured input-to-output mapping with preserved invariants generally, formalization specifies that the invariant targeted is the substantive content of the practice while the degree of freedom being reshaped is its representational explicitness — moving up from implicit know-how to codified knowledge.
Children (1) — more specific cases that build on this
-
Formal vs. Informal Structures presupposes Formalization
The formal-versus-informal duality requires that some practices be rendered explicit and codified into rules, charters, and policies — the very move that formalization names. Without the formalization machinery — making tacit practice statable, checkable, and transmissible — there would be no formal layer to contrast with the informal layer; the distinction would collapse into a single tier of uncodified practice. Formalization is the structural operation that produces the formal half of the duality, and is therefore presupposed by the dual structure.
Path to root: Formalization → Transformation
Neighborhood in Abstraction Space¶
Formalization sits among the more crowded primes in the catalog (13th percentile for distinctiveness): several abstractions describe nearly the same structure, so a description that fits it will tend to fit its neighbors too — transporting it usually means disambiguating within this family rather than landing on it exactly.
Family — Representation & Interpretive Mapping (25 primes)
Nearest neighbors
- Transformation — 0.84
- Institution — 0.83
- No One Is Above the Rules — 0.82
- Verification — 0.82
- Translation and Conceptual Bridging — 0.82
Computed from structural-signature embeddings · 2026-05-29
Not to Be Confused With¶
Formalization's closest and most easily confused neighbor is Emergent Formalization, and the two must be held sharply apart because the surface vocabulary is nearly identical while the structure is opposite on the dimension that matters most: agency and intent. Emergent Formalization names the unplanned, historical, often diachronic crystallization of regularity out of repeated use — the paradigm being grammaticalization, where a content word erodes over generations into a grammatical marker, or where a recurring usage pattern hardens into a rule that no one ever sat down to write. There is no author, no moment of decision, no draft; the "formalization" is an emergent statistical regularity that observers later discover and describe. Formalization, the present prime, is the reverse: a deliberate, authored act in which some agent, at an identifiable moment, decides a practice is worth codifying and undertakes the labor of articulating its elements and rules into an explicit, checkable artifact. The legislature codifying custom into statute is formalization; the slow drift of customary behavior into an unwritten norm that a sociologist later names is emergent formalization. The acid test is to ask who did it and when: if there is a drafter and a date, it is Formalization; if the regularity simply accreted through use and was only afterward recognized, it is Emergent Formalization. They can even chain — an emergent norm can later be deliberately codified — but the prime applies to the codifying act, not to the prior emergence. Confusing them is the single most important error to avoid here, because both can be described as "a practice becoming a rule," yet one is design and the other is drift.
Formalization is also distinct from Validation, the confirmation that a system meets its specification or requirements. The relation is sequential and complementary rather than overlapping: formalization produces the explicit system — the spec, the statute, the ruleset — and validation is a later activity that checks something against that explicit system. You cannot validate against an informal practice; the very thing that makes validation possible is that formalization has already supplied a statable standard to validate against. A team formalizes its release procedure into a runbook (formalization); an auditor later confirms that a particular release followed the runbook (validation). The two are frequently run by different people at different times, and a practice can be formalized and never validated, or repeatedly validated against a formalization that is itself wrong. Treating them as one move obscures the fact that formalization can succeed (a clean, checkable artifact exists) while validation fails (the practice does not in fact conform), and vice versa.
Finally, Formalization must be separated from Formal vs. Informal Structures, which is a descriptive prime naming the coexisting dual layers that organizations and systems exhibit — the documented org chart alongside the real influence network, the written policy alongside the way things actually get done. That prime is about the standing state of having both a formal and an informal layer, and the persistent gap between them. Formalization is the process that moves content from the informal layer into the formal one: it is the verb to that prime's noun. A consultant who observes that a company's real decision-making happens outside its documented hierarchy is invoking Formal vs. Informal Structures; a consultant who then writes the real process down as official procedure is performing Formalization. The distinction matters because the descriptive prime explains why formalization never fully succeeds — the informal layer regenerates as fast as the formal one is built — whereas the process prime names the specific, deliberate move that attempts the transfer. One names the landscape of two layers; the other names the act of pushing material from one layer to the other.
Solution Archetypes¶
No catalogued solution archetypes reference this prime yet.
Notes¶
Formalization operates across substrates that share a knowledge/practice character but differ sharply in their mechanics and reversibility. Logical and mathematical formalization is the cleanest case: the elements are symbols, the rules are inference rules, and the artifact is fully mechanical. Legal and organizational formalization is messier — the "rules" are statutes and procedures interpreted by humans, the conformance check is itself a judgment, and the artifact is continually amended. Computational formalization (specs, schemas, expert-system rulesets) sits between, mechanical in execution but authored under deep uncertainty about whether the encoded rules actually capture the tacit practice. The prime is the same move throughout, but a reasoner should not assume that the ease and fidelity of mathematical axiomatization carries over to the encoding of an expert's judgment.
The prime is largely absent from physical and biological substrates as a structural pattern, which is why its substrate-independence is scored at 4 rather than 5. Crystallization, ossification, and the hardening of a developmental pathway are sometimes described with formalization-flavored language, but these are metaphors borrowed from the knowledge case; nothing in a crystal "decides to codify" anything. The genuine instances of the prime all involve an agent acting on knowledge or practice, which keeps its breadth bounded to the social, formal, and computational domains.
A persistent confusion worth flagging is the conflation of formalization with rigor or quality. Formalization is necessary for certain kinds of rigor (mechanical checkability) but is neither sufficient for it nor identical to it: a precisely formalized system can encode nonsense with great rigor, and a deeply rigorous practice can remain entirely tacit in a master's hands. The prime names the move toward explicitness, not the achievement of correctness, and the most common practical error is to assume that having written something down means it is now right.
References¶
[1] Polanyi, M. (1966). The Tacit Dimension. Doubleday. Foundational statement that "we can know more than we can tell": a substantial residue of any skilled practice is tacit, which is why codifying it into explicit form transforms rather than transcribes it, and why the formal artifact is a lossy model of the practice it abstracts. ↩
[2] Hilbert, David. Grundlagen der Geometrie (Foundations of Geometry). Teubner, Leipzig, 1899. Develops axiomatic abstraction: geometry is abstracted into a set of axioms (incidence, betweenness, congruence, continuity) independent of intuitive geometric content. Demonstrates how abstraction of underlying assumptions makes geometry rigorous and reveals its true structure. ↩
[3] Nonaka, I., & Takeuchi, H. (1995). The Knowledge-Creating Company: How Japanese Companies Create the Dynamics of Innovation. Oxford University Press. Develops the deliberate conversion of tacit, person-bound know-how into explicit, codified, shareable knowledge (externalization) as an intentional, agent-driven organizational act; separates having a competent practice from having a stated, reusable, auditable system for it. ↩
[4] Frege, G. (1879). Begriffsschrift, eine der arithmetischen nachgebildete Formelsprache des reinen Denkens [Concept-Script: A Formal Language for Pure Thought Modeled on That of Arithmetic]. L. Nebert. Paradigm logical formalization: introduces a notation explicit enough (with quantification and the first modern predicate calculus) to make inference itself an object of inspection rather than an exercise of intuition. ↩
[5] Code civil des Français [French Civil Code / Napoleonic Code]. (1804). Imprimerie de la République, Paris. Codification of fragmented French customary law (droit coutumier), much of it oral, into a single explicit written statute, converting "what is usually done" into binding rule a citizen can read rather than infer from precedent. ↩
[6] Weber, M. (1922/1978). Economy and Society: An Outline of Interpretive Sociology (G. Roth & C. Wittich, Eds.). University of California Press. Foundational sociological theory: distinguishes rational-legal, traditional, and charismatic modes of legitimate domination, and ties modern adjudication to rule-bound rational-legal authority backed by the state's monopoly on legitimate violence. ↩
[7] Feigenbaum, E. A. (1984). Knowledge engineering: The applied side of artificial intelligence. Annals of the New York Academy of Sciences, 426(1), 91–107. Foundational knowledge-engineering account of encoding a domain expert's tacit know-how into machine-applicable rulesets and ontologies, naming the knowledge-acquisition bottleneck: experts cannot fully state what they know. ↩