Fact-Value Distinction¶

Origin domain: Philosophy
Subdomain: metaethics and decision governance → Philosophy
Also from: Law & Governance, Medicine & Healthcare, Engineering & Design, Science Policy
Aliases: Fact Value Dichotomy, Fact Value Gap, Descriptive Evaluative Distinction
Related primes: Normativity, Deductive Reasoning, Classification, Evidence

Core Idea¶

The fact-value distinction is the discipline of typing claims according to whether they describe what is or occurs, or evaluate what is good, justified, permissible, required, or worth pursuing. It prevents an evidential record from being read as though it already contained the standard by which the record should be judged. A description can constrain a choice, eliminate an option, expose a consequence, or show that a goal is infeasible; it does not silently create the value premise that makes one feasible outcome preferable to another.

The distinction is therefore more than the slogan that facts and values differ. It couples a two-class contrast to an explicit bridge requirement. When reasoning moves from a descriptive claim to an appraisal, ranking, entitlement, or recommendation, the analyst asks which value-bearing standard, reason, institutional mandate, or decision criterion licenses the move. If that bridge is absent, the conclusion has acquired evaluative force that the stated premises do not account for. If the bridge is present, facts and values can interact without being collapsed.

The prime does not require a metaphysical claim that facts and values occupy wholly separate worlds. It remains useful under moral realism, naturalism, constructivism, pragmatism, and value-laden accounts of inquiry. Those positions disagree about how facts and values are grounded and entangled. They can still distinguish the role of evidence from the role of a standard in a particular decision record and ask what authorizes movement between them.

Structural Signature¶

a descriptive claim class — an evaluative or normative claim class — a typing rule — an attempted crossing — a value-bearing bridge — facts that constrain without silently prescribing — an auditable conclusion

Descriptive class: claims about states, events, causes, probabilities, measurements, or observed regularities.
Evaluative class: claims that rank, commend, condemn, permit, require, justify, or supply reasons for action or belief.
Typing rule: a way to tell which role a claim plays in the argument, even when ordinary language mixes the two.
Attempted crossing: an inference, policy recommendation, appraisal, or design choice that moves from a factual record toward an evaluative verdict.
Value-bearing bridge: an explicit standard, goal, right, duty, authority, priority, or reason that gives the conclusion its evaluative force.
Constraining evidence: facts may change which actions are possible or which actions serve the standard without becoming the standard themselves.
Auditability: disagreements can be located in evidence, values, priorities, or the bridge instead of being hidden in the conclusion.

The signature is present only when both claim classes and their relation matter. A classification among two descriptive data types is not a fact-value distinction, and a free-standing norm with no descriptive contrast is simply normativity.

What It Is Not¶

Not a claim that facts are irrelevant to values. Evidence can determine consequences, feasibility, causal efficacy, and risk. Those facts can decisively alter what follows once a value premise is supplied.
Not a claim that values never shape facts. Measurement choices, categories, thresholds, and research agendas may be value-laden. The distinction asks that those roles be identified; it does not assume a view from nowhere.
Not moral subjectivism. A moral realist can hold that evaluative claims are objectively true while still distinguishing them from descriptive claims and requiring the relevant bridge in an argument.
Not the Is-Ought Problem. Is-Ought is the strict inferential specialization concerning deductive entailment from an all-descriptive premise set to a normative conclusion. The broader distinction also governs semantic typing, evidential authority, decision records, and institutional role separation.
Not Moore’s Naturalistic Fallacy. Moore’s thesis concerns whether “good” can be defined as a natural property. Fact-value distinction is broader and does not settle that definitional debate.
Not a strict MECE partition of every sentence. Mixed utterances, thick concepts, and institutionally loaded measurements can have both descriptive and evaluative functions. The prime requires the roles to be inspectable, not every sentence to occupy one context-free box.

Scope of Application¶

The distinction operates wherever an evidential record is used to justify an appraisal or action.

Ethics and metaethics: it separates empirical premises about consequences or human capacities from the principles that make those facts morally relevant.
Law and governance: it distinguishes what a rule currently authorizes from arguments about what the rule should authorize, while allowing evidence about effects to bear on reform.
Science-policy interfaces: assessment bodies can state what evidence supports and separately identify the risk tolerance, distributional priority, or mandate behind a recommendation.
Medicine and public health: outcome evidence, burden estimates, patient values, equity commitments, and acceptable-risk thresholds occupy different roles in a guideline.
Engineering and design: measured performance does not rank safety, cost, accessibility, and speed until design priorities or constraints are stated.
Journalism and institutional reporting: verified description, interpretation, endorsement, and decision criteria can be recorded separately.
Organizational governance: a decision log can distinguish what was known, what the organization valued, who had authority, and why the chosen action followed.

Its boundary is a practice capable of description and evaluation. A physical process by itself does not instantiate the distinction; an investigation or decision about that process can.

Clarity¶

The prime clarifies two errors that otherwise look like opposites. Value smuggling presents a recommendation as the inevitable output of neutral facts, concealing the criterion that selects it. Fact exclusion treats value judgment as insulated from evidence, as though causal consequences and feasibility could not defeat a proposed means. The distinction rejects both. It says that evidence and standards do different work and that sound reasoning often needs both.

This yields a practical diagnostic. Given a conclusion, label every load-bearing statement as descriptive, evaluative, or bridge-forming. Then ask whether changing the evaluative standard while holding the facts fixed could change the conclusion, and whether changing the facts while holding the standard fixed could do the same. If both counterfactuals are live, the decision genuinely depends on both kinds of input and should represent them separately.

Manages Complexity¶

Controversies at policy and design boundaries often entangle at least four disputes:

What happened or is likely to happen?
Which outcomes, rights, or obligations matter?
How should competing values be prioritized?
Does the proposed action actually serve those priorities under the evidence?

Without typed separation, a disagreement anywhere in that stack becomes a global fight over the final recommendation. The fact-value distinction decomposes the record so that empirical disagreement can be investigated, normative disagreement deliberated, priority conflict negotiated, and means-end fit tested. It turns one opaque controversy into several smaller and differently resolvable questions.

The same decomposition supports institutional division of labor without pretending the roles are independent. Analysts can own the factual model, authorized decision-makers can own the governing criteria, and the decision record can show the bridge between them. Accountability improves because neither expertise nor authority is allowed to masquerade as the other.

Abstract Reasoning¶

The prime licenses several reasoning moves.

Claim typing. Determine whether a statement reports a condition, supplies a standard, or connects a condition to a standard. Ordinary words are not decisive: “unsafe” may summarize a measured threshold fixed by regulation and therefore contain both descriptive and normative work.

Bridge extraction. Rewrite a fact-to-recommendation argument until the missing goal, right, duty, priority, or authority becomes an explicit premise. The repair may make the inference valid while relocating disagreement to that premise.

Paired counterfactuals. Hold the facts fixed and vary the value standard; then hold the standard fixed and vary the facts. This reveals which input controls which part of the verdict.

Standing analysis. Ask who is authorized to certify the facts, who is authorized to adopt the standard, and whether one actor has silently crossed from one role into the other.

Value-sensitive measurement. Inspect whether choices of variable, threshold, population, or time horizon already encode priorities. Such findings do not abolish the distinction; they locate evaluative content earlier in the pipeline.

Knowledge Transfer¶

A clinical guideline and an engineering design review share the same structure. The guideline records treatment effects and adverse-event probabilities, then combines them with patient values, equity commitments, and an acceptable-risk threshold. The design review records load tolerance, failure rates, cost, and latency, then combines them with safety margins and prioritized requirements. In both, measured facts constrain the available choice, an evaluative standard ranks the remaining options, and an explicit bridge explains the recommendation.

The transfer extends to law and public policy. An analyst may estimate that a rule reduces emissions, increases prices, and distributes burdens unevenly. None of those findings alone selects the rule. A decision follows only after commitments concerning climate harm, affordability, distribution, rights, and institutional mandate are introduced. The same audit learned in clinical evidence grading—separate effects, confidence, values, and recommendation—ports directly to the policy record.

Examples¶

Formal/abstract¶

Suppose a model supports the descriptive claims: action A reduces harm by ten units, costs two units, and transfers one unit of burden to group G. The conclusion “choose A” does not follow until the decision record supplies standards governing harm reduction, cost, and distribution. With a utilitarian aggregation rule the recommendation may favor A; with a side constraint forbidding the burden transfer it may reject A. The facts remain fixed while the value standard changes the verdict, showing that the standard is load-bearing rather than decorative.

Mapped back: the model outputs are the descriptive class, the aggregation rule or side constraint is the evaluative class, and the decision rule joining them is the explicit bridge.

Applied¶

A public-health report finds that a screening program detects disease earlier but produces false positives and unequal follow-up burdens. A recommendation to adopt the program requires judgments about acceptable false-positive harm, distributional equity, budget priority, and the authority under which those trade-offs are made. The report becomes more—not less—actionable when it displays the factual estimates and the value-bearing criteria separately, because later reviewers can update one without pretending the other changed.

Mapped back: measured effects and uncertainty are descriptive; acceptable harm and equity commitments are evaluative; the guideline method is the bridge that produces the recommendation.

Tensions¶

Transparency versus decision speed: fully exposing factual and evaluative premises improves auditability but increases deliberation cost.
Role separation versus entanglement: distinct institutional roles reduce hidden value smuggling, while measurement and framing choices can themselves contain values.
Neutral expertise versus responsible advice: experts should not disguise preferences as findings, yet refusing to explain the decision consequences of evidence can make expertise unusable.
Stable standards versus revisability: fixed criteria support consistent decisions, while new facts and moral learning may justify changing the criteria themselves.
Clear typing versus thick concepts: sharp categories aid audit, but concepts such as harm, disease, fairness, and risk often combine descriptive and evaluative content.

Structural–Framed Character¶

Fact-Value Distinction is a framed prime. Its relational skeleton—two claim classes plus a bridge—is abstract, but the distinction exists inside practices of giving reasons, adopting standards, and authorizing decisions. “Fact,” “value,” “ought,” and “justification” do not shed their evaluative meaning when the concept travels. They are imported intact into law, medicine, engineering, and policy.

Its framed character does not make it domain-specific. The same test is recognized across many institutional substrates: identify the empirical record, identify the governing standard, and expose the bridge. What stays constant is not a physical mechanism but a portable discipline for reason-giving practices.

Substrate Independence¶

The prime has composite substrate independence 4 / 5. Its domain breadth and transfer evidence are high: the exact distinction and bridge test recur in philosophy, law, medicine, science policy, engineering design, journalism, and organizational governance. Structural abstraction is slightly lower because the content classes remain evaluative and practice-bound rather than purely formal.

Composite substrate independence — 4 / 5
Domain breadth — 5 / 5
Structural abstraction — 4 / 5
Transfer evidence — 4 / 5

Relationships to Other Abstractions¶

Current abstraction Fact-Value Distinction Prime

Parents (1) — more general patterns this builds on

Fact-Value Distinction is part of Normativity Prime

Fact-Value Distinction contains Normativity because one side of the distinction must consist of genuinely evaluative, reason-giving, or ought-bearing content.

Children (1) — more specific cases that build on this

Is-Ought Problem Domain-specific is a kind of Fact-Value Distinction

The Is-Ought Problem is the Fact-Value Distinction specialized to deductive non-entailment from an all-descriptive premise set to a normative conclusion.

Hierarchy path (1) — routes to 1 parentless root

Fact-Value Distinction → Normativity → Constraint

Neighborhood in Abstraction Space¶

Fact-Value Distinction has no computed distinctiveness yet.

Family — Unclustered & Miscellaneous (429 primes)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-26

Not to Be Confused With¶

Is-Ought Problem — the deductive metaethical subtype. Tell: is the question specifically whether all-descriptive premises entail a normative conclusion, or more broadly how evidence and evaluative standards are typed and joined?
Normativity — the structure of standards and ought-bearing evaluation. It supplies one side of this distinction but not the descriptive contrast or bridge discipline.
Classification — the general assignment of items to categories. Fact-value distinction uses classification, but its identity lies in these two content roles and the governance of crossings between them.
Partition — a mutually exclusive and collectively exhaustive carve-up. The fact-value distinction need not claim that every utterance is context-free and purely factual or purely evaluative.
Naturalistic Fallacy — Moore’s definitional thesis about goodness and natural properties. The fact-value distinction does not by itself decide whether evaluative properties can be natural.
Value-free inquiry — a stronger thesis that values can or should be excluded from inquiry. This prime instead makes value-bearing choices explicit wherever they occur.

Solution Archetypes¶

No catalogued solution archetypes reference this prime yet.

Notes¶

This prime was authorized by the missing-node adjudication at workspace/mixed_dag_2026/missing_node_adjudications/fact_value_distinction.yaml. Its initial prose and examples require independent editorial and citation review.