Meta-Cognitive Reflection Prompt¶
Reflection prompt — instantiates Meta-Symbolic Rule Reflection
A short structured prompt that makes a reasoner name the categories and assumptions shaping their thinking before they apply the framework one more time.
A Meta-Cognitive Reflection Prompt is the lightest instantiation of the archetype: a short, structured question — asked at a decision point — that makes a reasoner pause and name the symbolic system they are reasoning inside before they apply it again. Its defining move is that it is a trigger, not a revision: it does no critique, proposes no new categories, and governs nothing; its entire job is to make the invisible framework momentarily visible to the person using it, in the seconds before habit re-applies it. It is the front door of the archetype — the thing that turns "just classify it the usual way" into "wait, which categories am I forcing this into, and what are they assuming?" — and it succeeds when it hands a now-conscious framework to a deeper mechanism, or fails safely by prompting nothing.
Example¶
An equity analyst is about to finalize a "sell" rating on a retailer whose numbers have slipped. Her desk has a standing rule: before any rating ships, answer three prompted questions on a card. What category am I putting this company in, and what does that category assume? What would I have to see to be wrong? Whose framing am I borrowing? The prompt makes her write it down: she has slotted the retailer as "declining brick-and-mortar," a category that assumes physical stores are a liability and treats an expanding store footprint as bad news. Naming the assumption, she notices the company is opening stores precisely because they double as fulfillment hubs — a fact the "declining retail" category renders invisible by default.
The prompt does not tell her the rating is wrong; it does not re-classify the company or revise the desk's category set. It simply surfaced the frame she was about to apply on autopilot, which is enough to send her to a real analysis of whether "declining brick-and-mortar" is the right box at all. Without the prompt she would have shipped the rating and never seen the assumption.
How it works¶
- Fire at the point of application. The prompt is placed exactly where a framework is about to be re-applied — before the rating ships, before the ticket is closed — not in a detached retrospective.
- Ask two things only. Which symbolic system am I in? and what is it assuming? Keeping it to naming-the-frame and naming-the-assumption is what makes it cheap enough to actually use every time.
- Externalize the answer. The reasoner writes the frame and its assumption down, because an unrecorded "yes I considered it" collapses back into autopilot instantly.
- Hand off, don't resolve. When the prompt surfaces a suspect frame, it routes to a deeper mechanism; it never tries to fix the framework itself.
Tuning parameters¶
- Trigger placement — every application, or only high-stakes ones. Universal placement catches more but risks becoming ritual noise; selective placement preserves salience.
- Prompt specificity — generic ("what am I assuming?") versus domain-tailored ("am I treating store count as a liability?"). Specific prompts bite harder but must be maintained as the domain shifts.
- Response demand — a checkbox, or a written sentence. Writing forces real reflection but adds friction that erodes compliance over time.
- Escalation rule — what a surfaced doubt triggers: a private note, or a mandatory hand-off to review. Stronger escalation acts on more doubts but makes people answer the prompt defensively.
When it helps, and when it misleads¶
Its strength is catching the frame you were about to apply unconsciously, at the one moment intervention is cheap — before the classification is committed. It is a deliberately minimal version of consider-the-opposite[n1], the debiasing move of forcing yourself to articulate the assumption and its alternative rather than sliding straight to the habitual answer.
Its central failure is that a prompt is not the archetype — surfacing an assumption changes nothing unless it leads to critique, revision, and governance, and a prompt answered by rote ("assumptions considered: yes") is worse than none, because it launders autopilot as reflection. Over-used, it becomes exactly this ritual: a box everyone ticks and no one reads. The classic misuse is treating the presence of the prompt as evidence that reflection happened. The guarding discipline is to demand a written, specific answer and to wire a real escalation path, so a surfaced doubt actually reaches a mechanism that can act on it rather than dying on the card.
How it implements the components¶
symbol_system_map— informally: the prompt makes the reasoner name, in the moment, which categories and rules they are about to apply, giving a one-line map of the frame in play.rule_assumption— its central ask is to state what that framework treats as normal, given, or impossible — surfacing the embedded assumption before it fires again.
It deliberately stops there: it runs no meta_level_critique (that is Category Boundary Audit or Paradigm Review), produces no revision_proposal (that is Model-Class Revision or Ontology Revision Session), and owns no validation_and_governance (that is Governance Rule Review).
Related¶
- Instantiates: Meta-Symbolic Rule Reflection — it is the archetype's trigger: the minimal prompt that makes a framework visible before it is re-applied.
- Sibling mechanisms: Category Boundary Audit · Paradigm Review · Language / Rule Critique · Edge-Case Walkthrough · Governance Rule Review · Model-Class Revision · Ontology Revision Session · Taxonomy Redesign Workshop · Versioned Schema Change
Editorial Notes¶
Form Classification¶
Form family: Interface, Display & Cue
Rationale: Meta-Cognitive Reflection Prompt operates as a user-facing prompt, display, template, or perceptual cue that shapes attention and action at the point of use because it a short structured prompt that makes a reasoner name the categories and assumptions shaping their thinking before they apply the framework one more time.
Independent corroboration: The frozen evidence defines Meta-Cognitive Reflection Prompt as 'A short structured prompt that makes a reasoner name the categories and assumptions shaping their thinking before they apply the framework one more time', so its operative form is Interface, Display & Cue.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Education & Pedagogy
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Prompts that make learners inspect their reasoning belong to metacognitive instructional practice.
Related originating lineages:
- Cognitive Science — Metarepresentation and strategy monitoring supply the information-processing account.
- Psychology — Self-regulation research shaped reflection on attention, confidence, and assumptions.
Review outcome: Independent reviewer agreement; high confidence.
Notes¶
The prompt is valuable precisely because it is thin — it is the only mechanism here cheap enough to run on every case, which is where autopilot lives. Loading it with critique or revision duties would make it too heavy to fire routinely, and it would stop catching the everyday framings that the weightier mechanisms never see because those cases never reach them.
[n1] Consider the opposite — a debiasing strategy (studied by Lord, Lepper, and Preston) in which a reasoner is prompted to explicitly ask what would make their current judgment wrong, which reliably reduces overconfidence and confirmation bias. The reflection prompt is a standing, structural version of the same move aimed at the reasoner's categories rather than their conclusion. ↩