Think Frame Creation

Frame Creation

Some problems do not yield to better solutions because the frame they arrived in is the obstacle. More police did not fix the late-night entertainment district; more generous-vs-limited dials do not settle a free tier. Frame creation refuses to solve inside the given frame. It explores the broader situation, distils its underlying themes, locates the core paradox and the value actually sought, then abduces a new working principle - a genuinely new standpoint, usually crystallised as an "approach it as if it were Y" reconception that redefines what the problem is - and only then reasons forward to the solution directions that frame unlocks. The reconception is the durable move; it is the de-branded core of Kees Dorst's frame creation (Dorst 2011; Frame Innovation, 2015). The output is a frame proposal, not a discussion, and never a proven answer.

When to Use

• The problem is genuinely open, complex, and paradoxical, and conventional problem solving inside the frame it arrived in has already failed or is producing more of the symptom.
• The way the problem is framed is itself the obstacle - the people closest to it have (mis)framed it, and the leverage is in re-seeing what kind of problem it is rather than optimising the current one.
• There is a real conflict of standpoints or requirements (a core paradox) that cannot be resolved head-on.
• A fresh standpoint would be worth far more than another solution attempt inside the existing terms.

When NOT to Use

• The problem is closed, familiar, or already well-framed. When the situation is settled and a frame comes to mind straight away, the elaborate move only manufactures a paradox that is not there - the same failure as reframing a correct problem. Frame creation earns its cost only when the situation presents a real paradox. This is the central wall.
• The reframe drifts off the real goal (goal-reformulation failure). The documented failure mode (Vermaas and Dorst, 2015) is reframing so freely that you end up solving the reframed problem, not the original one. The value actually sought has to anchor the frame, or the move produces a clever answer to the wrong question.
• The frame cannot be adopted by the people who own the problem (frame failure). A frame the client or stakeholders will not take up is inert, however elegant. This skill surfaces a new standpoint; it does not by itself make a powerful actor accept it. Check adoptability before proposing.
• It is mistaken for surface analogy or for solution generation. The "as if it were Y" line is the crystallisation of a frame built from themes, not a free-association prompt. Run as "pick a cool analogy and brainstorm" it collapses into ordinary ideation. For an analogy pointed at the solution with the problem held fixed, use think-far-analogy-ideation; for menu-driven rewordings of the given problem, use think-problem-restatement.
• A frame is treated as definitive before it is tested. The frame cannot be accepted as proven until a design built on it has been shown to deliver the value sought. The output is a promising standpoint to develop and test, never a settled answer.

Instructions

When asked to reframe an open, stuck problem, follow these steps:

1. Confirm it is an open, paradoxical problem. State the problem as given and check that solving inside its current frame has failed or is producing more of the symptom. If the problem is closed, familiar, or already well-framed, stop and say so - do not manufacture a paradox. If it is merely vague rather than genuinely two-sided-and-stuck, route to think-problem-restatement first.
2. Explore the broader context. Look around the problem, not only at it. Gather who is involved, what they are actually trying to do, the surrounding forces and history, and what conventional solutions inside the frame have already tried and failed. The new frame emerges from this engagement with the broader context, so do not skip to the analogy.
3. Distil the underlying themes. From the broader context, name the recurring underlying patterns - the experiences, motives, and meanings that actually drive the situation (a sense-making move, not a list of facts). These themes are the raw material the new frame is abduced from.
4. Locate the core paradox and the value actually sought. Name the real conflict that makes the problem hard (the standpoints or requirements that cannot both be satisfied head-on). Then name the value actually sought - what success would really deliver. Do not attack the paradox head-on; use it as the diagnostic that the current frame has failed, and let the value anchor everything downstream (this is the guard against goal-reformulation drift).
5. Abduce a new working principle - the reframe. From the themes, construct a genuinely new standpoint: a working principle that, if adopted, would let the value be created. Crystallise it as an "approach it as if it were Y" reconception that redefines what the problem is (the music-festival reframe of a late-night district; not "this is a crime problem" but "this is a hospitality problem"). Y must be earned by the themes, not free-associated, and it must change the problem, not just suggest a solution. State it as an IF/THEN implication: IF we see this situation as Y and adopt its working principle, THEN we create the value sought.
6. Reason forward to solution directions. Derive the solution directions the new frame generates natively - the needs and moves that become obvious once the problem is seen as Y (a festival frame generates staggered transport, wayfinding, friendly guides). These are derived forward from the frame, not transferred mechanisms from the source domain. Sanity-check adoptability: would the people who own the problem take up this standpoint?
7. Emit the frame proposal. Produce the artifact in references/TEMPLATE.md: the themes distilled, the core paradox and the value sought, the new frame as an IF/THEN "as if it were Y" implication, and the solution directions it unlocks - flagged explicitly as a standpoint to develop and test, not a proven answer.

Output Format

Use the template in references/TEMPLATE.md. The deliverable is the filled frame proposal - themes, core paradox and value, the IF/THEN "as if it were Y" frame, and the solution directions it unlocks - not a prose essay, and explicitly marked as untested.

Quality Checklist

Before finalizing, verify:

• The problem is genuinely open and paradoxical, and solving inside its current frame has failed - not a closed or already-well-framed problem with a manufactured paradox.
• The frame was built from distilled themes of the broader context, not grabbed as a surface analogy or free-associated.
• The core paradox is named, and the value actually sought is named and anchors the frame (guard against goal-reformulation drift).
• The new frame redefines what the problem is (an "as if it were Y" reconception stated as an IF/THEN implication), not a reworded problem and not a solution analogy under a fixed frame.
• The solution directions are derived forward from the new frame (what it generates natively), not mechanisms transferred from the source domain.
• Adoptability was checked - the frame is one the people who own the problem could actually take up.
• The output is the frame-proposal artifact, presented as a standpoint to develop and test, never a proven answer.
• No overclaiming: the evidence is conceptual and transferred from human design practice; claim "constructs a new, theme-grounded standpoint that generates native solution directions," not a measured improvement in outcomes (see evidence/dossier.md).

Evidence

Tier C (governing; honest read C/P, capped at C). Frame creation is an influential, well-developed account of design reasoning, built on roughly fifty years of design research and Dorst's own observational studies of expert designers (Dorst 2011; Frame Innovation, 2015), with worked case studies from the Designing Out Crime centre; the general proposition that how a problem is framed shapes the solutions found has moderate support in the wider problem-framing literature (Schon's reflective practice; problem-finding research; small-N design-team studies). But there is no controlled, comparative, or outcome study of the named method - no trial showing the move produces better problems, frames, or results than not running it; the cases are demonstrations, not measured comparisons with a baseline or control. The frequently-cited "about 20 design directions from one frame" is an illustrative count from a single case, not an effect size, and does not influence the tier. The entire base is human designers and human organisations, transferred and not validated for AI-augmented use - a second reason the grade is C. Full grading, sources, and caveats: evidence/dossier.md.

Examples

See references/EXAMPLE.md for a completed frame proposal on a real decision.