Everything is bold, so nothing is

A lesson from UI and UX for Engineers: Decisions You Made By Accident. Play it above, or read it through below.

• Emphasis is a property of the crowd.

  The common read: If a reviewer can see the difference between two colours, users will spot the one that matters in a full screen.

  Pop-out is not a property of a colour. It is a property of a target inside a particular display, and the display is the half a design review never contains. Search efficiency rises with the difference between the target and the distractors and falls with the difference among the distractors themselves, which means the same accent colour can be found instantly on one screen and be invisible on the next.

  There is a second condition, and it is the one with real design consequences. Search is efficient when the target is linearly separable from the distractors in the feature space, which means a line can be drawn with the target on one side and every distractor on the other. When the distractors flank the target, search collapses. An accent that sits between two other accents you already ship is the flanked case, and no amount of saturation rescues it.

  The thresholds are also much coarser than intuition. For orientation, the reported critical difference for efficient search is around fifteen degrees, far larger than the smallest tilt a person can detect, and guidance turns out to be categorical rather than metric: search is more efficient when the target is uniquely steep or shallow. Colour behaves the same way, and the researchers say so directly: the metrics of colour difference for foveal discrimination are quite unlike those that govern the deployment of attention.

  That arrives as a ticket titled "make the overdue state more obvious", assigned to whoever owns the palette, closed by a hex change, and reopened next quarter with the word "still". Count your accents before you add one: the question is not whether the new colour is different enough from the background, it is whether it is different enough from every other thing on that screen already competing to be noticed.

• The F is the shape of a page that gave up.

  The common read: People scan web pages in an F, so the design job is to put what matters into the top band and down the left edge.

  The F is what the eye does when you have given it nothing, and the group that found it now says so in those terms. Their sentence, quoted whole: "The F-shaped scanning pattern is bad for users and businesses: it means that users may skip important content simply because it appears on the right side of the page." Designing for the F is designing for your own failure.

  It also has preconditions, and they were in the original. It appears when the text has little or no formatting for the web, when the user is trying to be efficient, and when the user is not committed enough to read every word. Remove any one of those and the pattern changes. The mechanism is stated plainly too: in the absence of signals to guide the eye, people take the path of minimum effort and fixate near where they started reading.

  There are other patterns and they are ranked. Spotted scanning lands on scattered words. Layer-cake scanning lands mostly on headings and subheadings, and is described as by far the most effective way to scan. Commitment scanning is actual reading. Worst to best, the order is F, spotted, layer cake, commitment.

  That arrives as an analytics review where a long feature page has a scroll depth of thirty percent, filed as a content-length problem, answered by cutting the page in half, which changes nothing because the surviving half is still four unbroken paragraphs. The lever is a typographic one and it is unglamorous: write subheadings that carry information rather than category names, and the pattern you get is the good one.

• Two of these were measured. The rest is craft, and craft is fine.

  The common read: Size, weight, position and isolation are the four measured levers of visual hierarchy, and a designer's hierarchy model predicts where the eye lands.

  Two things in the hierarchy vocabulary have been measured and the rest is craft, and the honest version of this beat is that the craft is still worth learning. The 2018 test of a published hierarchy model found it predicted entry points poorly and was mainly selecting on element size alone. What survived was spatial position and distinctiveness, and its conclusion tells designers to stop using the model because its predictions can be misleading.

  The measured vocabulary and the design vocabulary do not line up, and that is worth seeing directly. The canonical computational attention model builds forty-two feature maps from three channels: intensity contrast, colour opponency and orientation contrast. Size, weight, position and isolation appear in none of them. The nearest thing to isolation in that model is a normalisation step that promotes a map when its most active location is far above the map's own average, which is feature uniqueness within a channel rather than whitespace around a box.

  The design vocabulary is not therefore worthless. It is a teaching tradition with no experimental literature behind its specific enumeration, and even the most research-forward usability shop presents it citing two design textbooks and no experiments. That is not a scandal. It is the actual epistemic status, and knowing it changes how you argue: you can defend a hierarchy decision as a craft judgement and you cannot defend it as a finding.

  That arrives as a design review that stalls for forty minutes, where two competent people cite incompatible principles at each other and the outcome is decided by seniority, then gets written into a component library as a rule. Argue the two measured things when you have them, put the rest in the same sentence as the words "my judgement", and spend the review time you save on the one question the literature does answer: what else on this screen is competing.

  Make one thing loud. Then make everything else quiet, and quiet the same way. Loud thing must be far from quiet things. Not a little far. Far enough that a line fits between. Many different loud things? Then no loud thing. Only noise. Rule: emphasis is about the crowd. Never about the colour.

• Isolation has a name, and the name is about memory.

  The common read: Isolating an element is what makes it noticeable, and the isolation effect is the research that says so.

  What does the isolation effect actually predict? Not that you will find the odd item faster. The Von Restorff effect is a MEMORY result: an item that stands out from its surroundings is better remembered. Recall and search are different measures, and the design literature routinely cites the memory finding to support an attention claim.

  That distinction has a practical edge rather than being pedantry. If you isolate the destructive action in a menu, you have a memory result on your side and an attention result you have not established. Both may be true. Only one of them has the study everyone is citing, and if the argument you are making is about whether users will FIND the control, the isolation effect is not the citation you want.

  Whitespace is the other half of the confusion. Space around an element is real and does work, and what it does is grouping, which the next lesson takes as its subject. It is not the normalisation term in a saliency model, and it is not a fifth guiding attribute. Isolation as designers use the word bundles a grouping effect and a memory effect and a folk theory of attention into one instruction.

  That arrives as a component-library pull request that adds twenty-four pixels of margin above a delete button and cites the isolation effect in its description, approved by two reviewers who recognise the term. Keep the move and change the argument: the space is there because the button must not read as part of the group above it, which is a grouping claim you can defend, and the memory result is a bonus you did not need to invoke.

• Emphasis is one input. Nobody published the exchange rate.

  The common read: A strong enough hierarchy carries a layout, so moving a familiar control is safe as long as the new position is visually prominent.

  Which wins, a high-contrast element in an unusual place or a low-contrast element where users expect it? Nobody has published the answer, and that absence is the honest close to a lesson about emphasis. The current model of visual search names five sources of guidance, and features are one of them. The learned structure of a scene, its syntax and its semantics, is another named source sitting beside them.

  Read what that does and does not license. It licenses saying that emphasis is one input among several, and that where a thing usually lives is a different input rather than a weaker version of the same one. It does not license a ranking, because none was measured, and a course that told you convention beats contrast would be inventing an exchange rate to replace the one it just spent four screens refusing.

  The practical consequence survives without the ranking. A redesign that relocates a familiar control changes an input the model names, and there is no published rate at which the feature input buys it back. So the compensation everybody reaches for, making the moved thing louder, is a trade nobody can price, which means it is a bet rather than a plan.

  That arrives as a support queue full of "where did search go", filed in the week after a launch, answered with a help-centre article, and closed when the volume drops because the people who could not find it stopped looking. Move a familiar thing when the reason is worth an unpriced cost, say out loud that it is unpriced, and measure the move rather than arguing it, which is what the next level is about.