Domain Restriction¶
Protocol — instantiates Solution Space Bounding
Restricts the candidate space to a chosen category — a jurisdiction, market, population, technology family, or discipline — argued to be where the relevant answers live.
Domain Restriction bounds the space by membership in a category: it keeps only candidates that belong to one field, family, market, or class and excludes all others by kind, wholesale. The single idea that makes it this mechanism is that the boundary is qualitative and argued — "the answer lives in this domain, so we search there" — rather than numeric (a cap) or temporal (a window). Its signature liability follows directly from that move: because it excludes entire categories at once, it can amputate a valuable cross-domain candidate in a single stroke, which is why a legitimate domain restriction carries an explicit relevance argument and a review aimed squarely at that amputation risk.
Example¶
A solar-cell research group faces an effectively open-ended space of candidate photovoltaic materials — far too many chemistries to synthesize and test. It restricts its program to the perovskite family, on a recorded argument: perovskites have shown unusually rapid efficiency gains and cheap solution processing, so the domain is where near-term payoff is most likely. Everything non-perovskite is out of scope for this program. Crucially, the team runs a false-exclusion review and flags an adjacency it is not willing to lose blind — silicon–perovskite tandem cells straddle the boundary — and keeps that hybrid class under explicit watch. The search is now tractable within one family, and the one cross-domain candidate that could have been the biggest win is named rather than silently discarded.
How it works¶
The protocol fires on a specific signal: candidates sprawl across too many domains to search them all. It then selects a domain by relevance argument (why this field, market, or family is where the decision's answer most plausibly lies), writes the membership rule that decides what is in-kind, and — because categorical exclusion is coarse — runs a false-exclusion review sized to the stakes, focused on high-value candidates just across the boundary. What separates it from a record-level filter is that it excludes by category, not by attribute: it does not ask "does this item pass a threshold?" but "is this item the kind of thing we are considering at all?"
Tuning parameters¶
- Domain breadth — a narrow domain is more tractable and comparable but loses more cross-domain novelty; a broad one hedges novelty at the cost of search feasibility.
- Number of domains admitted — a single family versus a few adjacent ones. Admitting adjacencies softens the amputation risk but reintroduces heterogeneity.
- Relevance-argument strictness — how demanding the justification must be before a domain is accepted as the frame.
- Cross-domain review rigor — how hard the false-exclusion review looks for valuable out-of-domain candidates; scale it to reversibility and stakes.
When it helps, and when it misleads¶
Its strength is that restricting to one domain makes comparison apples-to-apples and search feasible: candidates within a family share units, methods, and evaluation criteria, so they can actually be weighed against each other.
Its failure mode is structural exclusion of the unconventional — and the biggest wins in many fields come from an adjacent domain the restriction quietly walled off. The classic misuse is the streetlight effect[1] — searching the familiar, well-lit domain because it is comfortable and legible, not because the answer is there. The guarding discipline is the false-exclusion review doing real work — naming the specific high-value adjacencies excluded and keeping them reviewable — plus a willingness to widen or shift the domain when evidence shows the answer sits outside it.
How it implements the components¶
possibility_space_signal— it is triggered precisely by the signal that candidates span too many domains to inspect exhaustively, warranting a categorical bound.boundary_rationale— the domain choice is a documented relevance argument, the thing that makes the categorical exclusion legitimate rather than arbitrary.false_exclusion_review— its built-in guard: a review that hunts for valuable cross-domain candidates the restriction may have amputated.
It does not count or cap candidates (cardinality_estimate, search_strategy) — that numeric ceiling is Candidate Cap, which bounds by how many rather than by kind; and it does not set a temporal window (scope_boundary, reopening_trigger) — that is Bounded Planning Window.
Related¶
- Instantiates: Solution Space Bounding — bounds the candidate space by category before search proceeds.
- Sibling mechanisms: Bounded Planning Window · Candidate Cap · Eligibility Screen · Finite Horizon Assumption · Sampling Frame Definition · Scope Statement · Search Filter
Editorial Notes¶
Form Classification¶
Form family: Decision, Gate & Allocation
Rationale: Domain Restriction operates as a case-specific gate, selection, routing, prioritization, or resource disposition because it restricts the candidate space to a chosen category — a jurisdiction, market, population, technology family, or discipline — argued to be where the relevant answers live.
Independent corroboration: The frozen evidence defines Domain Restriction as 'Restricts the candidate space to a chosen category — a jurisdiction, market, population, technology family, or discipline — argued to be where the relevant answers live', so its operative form is Decision, Gate & Allocation.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Operations Research
Origin pattern: Single lineage
Present-day reach: Multi-domain
Rationale: Formal problem solving cohered restricting a search to a justified feasible subdomain so effort is concentrated where admissible answers can occur.
Related originating lineages:
- Philosophy — Logic and philosophy of inquiry supplied explicit discourse universes and cautions against question-begging scope choices.
Review resolution: Formal problem solving supplies the clearest operational lineage, while logic supplies discourse-domain discipline; multi-domain reach is more calibrated than universal for this explicit search procedure.
Attribution caveat: Domain restriction is a very old formal move rather than a method with a unique modern disciplinary founder.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; medium confidence.
References¶
[1] Kaplan, A. The Conduct of Inquiry: Methodology for Behavioral Science. Chandler Publishing Company (1964). Warns against searching only where inquiry is easiest and most illuminated rather than where the answer is likely to be found. registry ↩