← The Casebook
Classified · cognitive-bias fileFile No. functional-fixednessDeclassify on: read

Functional Fixedness

Functional fixedness is the tendency to perceive an object only in terms of its familiar, designed function, which blocks solving problems that require using the object in a novel way.

The assessment

First demonstrated by Gestalt psychologist Karl Duncker (published 1945), functional fixedness is shown in tasks like the candle problem, where people fail to use a tack box as a candle shelf because it was primed as a container. The effect was replicated and tightened by Adamson (1952), who found people given an empty box solved roughly twice as often as those given a filled box, and it appears even among the technologically sparse Shuar of Amazonian Ecuador (German & Barrett, 2005) and in chimpanzees (Ebel et al., 2024). It is a robust, well-replicated phenomenon, though it overlaps closely with the Einstellung (mental set) effect and debates remain about whether it reflects object representation in semantic memory or problem-solving set.

Confidence in this file: strong grounding

What naive judgment can’t see

The dominant account is representational: an object's design/intended function is stored as a core property of its concept in semantic memory, so when that function is salient or primed, it pre-empts retrieval of alternative uses needed to solve the problem. Priming the typical function (pre-utilization) deepens the fixation; making the object's function less salient or decomposing it into generic parts (McCaffrey's technique) relieves it.

A rival/overlapping view treats functional fixedness as a special case of problem-solving set (Einstellung) — the carry-over of recent or habitual problem approaches — rather than a distinct property of object representation. Some work (e.g., contextual-dependency accounts) argues set and fixedness dissociate because certain variables affect them differently, implying partly separate mechanisms. The cross-cultural (German & Barrett) and animal (Ebel et al.) findings are used to argue the representational, possibly universal, basis.

Findings on file

  1. Duncker candle/box problem. Participants fixated on the box's container function and rarely used it as a candle holder when it was pre-filled; emptying the box improved solving. Karl Duncker, 1945

    effect: not reported (original monograph numbers not independently retrieved) · not reported

  2. Adamson replication of three Duncker experiments. Confirmed Duncker: pre-utilization of solution objects produced fixedness. Participants given an empty tack box were about twice as likely to solve as those given the box used as a container. Robert E. Adamson, 1952

    effect: approx. 2x higher solution rate with empty vs. filled box (as summarized; exact percentages not independently retrieved) · 57 participants (per secondary summary; Adamson's actual designs used separate Ns per problem)

  3. Birch & Rabinowitz two-cord (Maier) problem. Prior experience determined which object participants avoided using as the weight: they tended not to repurpose the object they had just used for its trained electrical function. Herbert G. Birch & Herbert S. Rabinowitz, 1951

    effect: not reported · not reported

  4. Glucksberg drive/incentive and fixedness. High incentive worsened performance when the task required overcoming fixedness (box-as-container) but helped when fixedness was absent (empty box); incentive narrowed focus and deepened fixedness. Sam Glucksberg, 1962

    effect: high-incentive group averaged about 3.5 minutes slower on the fixedness version (as summarized in secondary sources) · 128 male undergraduates (per secondary summary)

  5. German & Barrett, technologically sparse culture. Shuar participants showed functional fixedness (slower/worse solving when the object's typical function was primed), suggesting design-as-core-property of artifact concepts is not dependent on a tool-rich culture. Tim P. German & H. Clark Barrett, 2005

    effect: not reported (exact latencies/percentages not independently retrieved) · adolescents, mean age ~16 (Shuar) vs. industrialized controls

  6. McCaffrey Generic Parts Technique. Training reduced fixedness; trained participants solved substantially more problems than controls. Tony McCaffrey, 2012

    effect: trained group solved 67% more of the problems than controls · not reported in retrieved sources

  7. Functional fixedness in chimpanzees. After using a tool for one function, chimpanzees were less efficient at repurposing it for a new function; the effect persisted up to nine months and after learning the function by observing a conspecific. Sonja J. Ebel, Christoph J. Völter, Alejandro Sánchez-Amaro, Katharina A. Helming, Esther Herrmann & Josep Call, 2024

    effect: not reported (numeric) · not reported in retrieved summary

Verdict — replication

robust

The core effect has been independently reproduced repeatedly since Duncker: Adamson (1952) confirmed it with a larger, more controlled sample; Birch & Rabinowitz (1951) and Glucksberg (1962) extended it; German & Barrett (2005) found it in a technologically sparse culture; and Ebel et al. (2024) demonstrated it in chimpanzees. The phenomenon is generally treated as robust, though much of the classic work predates modern effect-size reporting and pre-registration, and many widely cited percentages come from secondary summaries rather than the primary papers. Its boundary with the Einstellung/mental-set effect is genuinely debated.

Recommended action

Use McCaffrey's Generic Parts Technique: decompose each object/resource into its parts and describe each in function-neutral terms (a 'candle' becomes 'a cylinder of wax with a wick of string'), which surfaces obscure features and uses. More broadly, separate an object's name/intended role from its raw properties before brainstorming solutions, and avoid priming the typical function.

McCaffrey (2012, Psychological Science) reported participants trained in the generic-parts technique solved 67% more functional-fixedness problems than untrained controls; Adamson (1952) and Glucksberg (1962) show that removing the priming of an object's typical function increases solution rates.

On file from Karl Duncker, 1945 — On problem-solving

Cross-referenced files

File your own case

Open the same case on your own draft.

Paste a memo, a research draft, or a strategy argument. It is scored against all 175 cards, and the strongest two or three risks come back with the evidence quoted and one practical next check.

Open a case on your draft →