Functional Fixedness in Design: The Candle Problem, Your Users, and Your Team

Hand someone a candle, a book of matches and a cardboard box full of thumbtacks. Ask them to fix the candle to a corkboard wall so it can burn without dripping wax on the table. Most people try to tack the candle straight to the wall, or melt its side and press it on. Both fail. The working answer is to tip the tacks out, pin the empty box to the wall, and stand the candle in it.

Karl Duncker published the task in 1945, and the interesting part was a second version of it. When the tacks were handed over loose, next to an empty box, far more people found the shelf (Duncker, 1945). Robert Adamson repeated the comparison in 1952 and got roughly twice the solve rate in the empty-box condition. The box was the same box. Presented as a container, it stayed a container.

That is functional fixedness: once you have seen an object doing one job, you have trouble seeing it do another. It belongs with the other biases in the UX psychology hub, but it behaves differently from most of them. Anchoring or the availability heuristic bend a judgment. Functional fixedness removes options before any judgment happens, so the person never notices they were missing.

What the classic experiments show

Three results carry most of the weight.

The candle problem (Duncker, 1945) shows that the way an object is first presented sets its function. Filled with tacks, the box reads as packaging and drops out of the list of available materials.

The two-string problem (Norman Maier, 1931) hangs two cords from a ceiling, too far apart to hold both at once, with a few tools on a table. The solution is to tie a pair of pliers to one cord and swing it as a pendulum. Herbert Birch and H. S. Rabinowitz later added a warm-up task: participants who had just wired an electrical switch into a circuit avoided the switch as their pendulum weight and picked the other object instead (Birch & Rabinowitz, 1951).

The age reversal (German and Defeyter, 2000) found that five-year-olds are less fixed than six- and seven-year-olds on a version of the box task (German & Defeyter, 2000). Younger children haven't yet learned that objects have a single intended purpose. Fixedness comes with knowledge, which is why experts are not immune to it and are often worse.

How it hides features from users

Every interface element gets introduced once, and that first introduction sets its function in the user's head. A star icon learned as "favorite" stays a favorite. It is rarely read as "add to the list that syncs to my phone," even if that is also what it does. A search box learned as search is not tried as a command line. A shared link learned as "send this to a colleague" is not thought of as a way to hand out a template.

The pattern shows up in research sessions as a specific kind of failure. The participant needs a capability, the capability is on screen, and they look straight at the control that provides it without trying it. When asked afterwards, they describe the control by its first job. They found the element and filed it under a different purpose, so the usual discoverability fixes (make it bigger, move it up) don't reach the problem.

Some practical consequences:

• Multi-purpose controls carry their first use. If one button does two things depending on context, most users will only ever know the first thing it did for them. Give the second job its own entry point, even a duplicated one.
• The empty state is where a second function gets introduced. A box already full of tacks can't be a shelf. A list screen with no items yet is the moment a user will read its purpose from its label and its prompt, before habit fixes it.
• Labels describe the job, not the object. "Folder" describes a container. "Share with a team" describes what it's for. When the object has more than one use, name the use you want people to find, at the point where they need it.
• Onboarding tours fix functions early. A tour that introduces a feature through its narrowest use case (the calendar is for meetings) makes the wider uses (the calendar is for deadlines, for booking rooms) harder to see later. Introduce the general function first.

When usability testing turns up a feature nobody uses, check whether participants can describe what the related control does before redesigning the feature. If they describe it confidently and wrongly, the problem is fixed function, and moving the control won't fix it.

Design fixation in the team

The same effect runs inside design teams, where it has its own name. David Jansson and Steven Smith asked engineering students to design products, including a spill-proof coffee cup, and gave some of them an example design first. The students who saw the example reproduced its features, including a straw and mouthpiece that the brief had pointed out as a drawback (Jansson & Smith, 1991).

For interface work this is a daily condition. Pattern galleries, competitor screenshots and teardowns are the example design in Jansson and Smith's study. Looking at twenty checkout flows before sketching one tends to produce the twenty-first.

• Sketch before you browse. Put down a few rough options for the problem first, then look at how others solved it. The references then check your ideas instead of replacing them. Lo-fi wireframes are cheap enough for this.
• Describe the problem without naming the object. "Users need a way to see what changed since their last visit" leaves more room than "we need a notifications panel."
• Bring in someone without the history. The people who built the current version know what each part is for, which is the knowledge that fixes function. A designer from another team, or a new hire in their first month, will suggest uses the owners have filtered out.

The ideation stage of design thinking exists partly to push against design fixation, but a workshop alone doesn't do it if the room starts by reviewing the existing product.

The generic parts technique

Most advice about functional fixedness is general ("think outside the box"). One method has a direct test behind it. Tony McCaffrey trained participants to break an object into its parts and describe each part without referring to its use: not "a candle" but "a cylinder of wax around a string." The string, described that way, can be pulled out and used as a tie. Trained participants solved substantially more insight problems than untrained ones (McCaffrey, 2012).

Applied to a product, the technique means listing what a feature is made of rather than what it does:

Feature, named by use	Its parts, named by form	Other uses the parts suggest
Comments on a document	Timestamped text attached to a position, with an author	Change log, review checklist, approvals
Saved searches	A stored query with a name	Alerts, shared views, onboarding presets
Favorites	A per-user list of object IDs	Offline sync list, recommendation seed

The third column is a list of candidates to test. Each one still needs evidence from users that the need exists.

Functional fixedness and related biases

Functional fixedness is often confused with a few neighbors:

• Mental set is sticking with a method that worked before. Functional fixedness is sticking with an object's first use.
• Confirmation bias is favoring evidence for what you already believe. It keeps a team from questioning a fixed function once one is in place.
• Anchoring is over-weighting the first number or option you see. The first presentation of an object works like an anchor for its purpose.

Frequently asked questions

What is functional fixedness in psychology?

A cognitive bias in which a person can only see an object in the way it is traditionally used or was first shown to them, which stops them from using it to solve a problem. The term comes from Karl Duncker's problem-solving research, published in 1945.

What is an example of functional fixedness?

The candle problem: people asked to mount a candle on a wall miss that the box holding the thumbtacks can be pinned up as a shelf. In a product, a user who has only ever used a star icon to mark favorites and never realizes the same list syncs for offline use is showing functional fixedness.

How do you overcome functional fixedness?

Describe objects by their parts and materials instead of their purpose (the generic parts technique), present objects without their usual context, and bring in people who haven't learned the "correct" use. For design teams, sketch options before looking at existing solutions.

What is the difference between functional fixedness and mental set?

Mental set is a fixed way of approaching problems, such as reusing a method that worked last time. Functional fixedness is a fixed idea of what an object is for. Both are forms of fixation, and they often occur together.

Is functional fixedness always bad?

No. Knowing what things are for is what lets people use tools quickly. It only becomes a problem when a solution needs an object used in an unfamiliar way, or when a product depends on users finding a second use for a control.

What is design fixation?

The version of functional fixedness that affects designers: seeing an existing example narrows the ideas a designer produces, often reproducing the example's features, flaws included. Jansson and Smith described it in 1991.

Sources: Karl Duncker, "On problem-solving," Psychological Monographs 58(5), 1945 · Robert E. Adamson, "Functional fixedness as related to problem solving," Journal of Experimental Psychology 44(4), 1952 · Norman R. F. Maier, "Reasoning in humans II: The solution of a problem and its appearance in consciousness," Journal of Comparative Psychology 12(2), 1931 · Herbert G. Birch & H. S. Rabinowitz, "The negative effect of previous experience on productive thinking," Journal of Experimental Psychology 41(2), 1951 · Tim P. German & Margaret A. Defeyter, "Immunity to functional fixedness in young children," Psychonomic Bulletin & Review 7(4), 2000 · David G. Jansson & Steven M. Smith, "Design fixation," Design Studies 12(1), 1991 · Tony McCaffrey, "Innovation relies on the obscure," Psychological Science 23(3), 2012.