The Raskin Center

The Humane Interface — Summary

“The Humane Interface: New Directions for Designing Interactive Systems” by Jef Raskin (Addison-Wesley, 2000)

Jef Raskin spent decades thinking about the relationship between computers and the humans who use them. The Humane Interface is the distillation of those decades — a systematic argument for why most computer interfaces fail users at a fundamental level, and a framework for designing interfaces that do not.

The book is not a style guide or a collection of usability tips. It is a theoretical account of how human cognition works and what that implies for interface design. Raskin was a cognitive scientist before he was a computer scientist, and it shows. The principles in this book derive from research in psychology and human factors, not from design intuition or aesthetic preference.

The Central Argument

The book’s core claim is direct: computer interfaces should be designed around how human minds actually work, not around how computers work internally. This sounds obvious. Raskin’s contribution was to show, rigorously and with specific examples, how far most interfaces are from achieving it — and to propose concrete principles for getting closer.

Raskin observed that most interface failures are not bugs. They are predictable consequences of design decisions that violate known properties of human cognition. Modal interfaces cause errors not because users are careless but because modes violate the way humans process information. Inconsistent behaviors impose cognitive load not because users lack practice but because inconsistency demands conscious attention that would otherwise be available for the task at hand. These are systematic failures with systematic causes, and they require systematic solutions.

Key Concepts

Habituation and Automaticity

One of the most important properties of human cognition for interface designers is the tendency for practiced actions to become automatic. When we learn to type, we begin by consciously selecting each key. With practice, the action becomes habitual — we think about words, not keystrokes. The cognitive load of the motor task drops to near zero.

This automaticity is enormously valuable. It is also a trap if the interface changes. An action performed automatically, without conscious attention, will be performed the same way even if the system has changed. If a menu item moves, a habituated user will click where it used to be. If a keyboard shortcut changes meaning in a different context, the habituated user will perform the wrong action without noticing.

Good interfaces leverage automaticity by maintaining consistency. Bad interfaces break it, typically by changing behaviors across contexts or versions, forcing users to remain conscious of details that should have become automatic.

Locus of Attention and Cognitive Load

Humans can consciously attend to one thing at a time. This is not a limitation to be overcome through better training; it is a fundamental property of the cognitive system. Interface design that ignores it — by requiring users to monitor multiple changing states simultaneously, or by interrupting a task with irrelevant information — imposes costs that cannot be avoided.

Raskin used the concept of “locus of attention” to describe where a user’s conscious focus is directed at any moment. Good interface design keeps the locus on the user’s task, not on the interface itself. Every time the interface demands attention — a dialog box, an error message, a mode indicator — it pulls the locus away from the task. This is sometimes unavoidable, but it should be recognized as a cost.

Modes Are Harmful

This is Raskin’s most influential argument, and the one most cited in subsequent HCI literature. A mode is a state of an interface in which the same user action produces different results. The Caps Lock key is the canonical example: in caps-lock mode, pressing a letter key produces a capital; in the default mode, it produces a lowercase. Same key, different results, depending on an invisible system state.

Modes are harmful because they create errors. Users lose track of what mode the system is in — especially after interruptions or distractions — and perform actions that produce unexpected results. The user typed what they intended. The system executed it correctly. The result is still wrong, because the user and the system had different models of the current state.

Raskin argued that modes should be eliminated wherever possible, or if unavoidable, made as visible and unambiguous as possible. He distinguished between modes, which are harmful, and quasimodes, which are not. A quasimode requires continuous action to maintain — like holding the shift key while typing. Because you must actively maintain a quasimode, you cannot accidentally forget you are in it. Quasimodes do not cause the class of errors that modes cause.

Quantifying Interface Efficiency

Raskin proposed that interface efficiency be measured rather than estimated. He drew on GOMS analysis and Fitts’s law to provide quantitative predictions of how long specific interface designs would take to use. This approach allows interfaces to be compared objectively rather than argued about on the basis of preference.

This was, and remains, a significant contribution. Interface design decisions are routinely made on the basis of what looks good, what the designer finds intuitive, or what the executive prefers. Raskin argued — and demonstrated — that many of these decisions can be subjected to empirical testing, and that the results sometimes contradict intuition. An interface that feels natural to an experienced user may be slower for a novice than one that feels awkward. Measurement resolves the dispute.

The Monotony of Good Design

Raskin made a point that designers often resist: good interface design is not visible. The best interfaces do not call attention to themselves. They allow users to focus entirely on their tasks. An interface that users praise for its elegance is usually one that stayed out of the way.

This has implications for how interface design is evaluated and rewarded. If the success of a design is measured by how little it is noticed, the people who design it receive less recognition than the people who design something conspicuous. Raskin thought this was a structural problem in how interface work gets valued, and said so.

The Archy Implementation

Raskin was not interested in purely theoretical contributions. He wanted to see these principles implemented in working software. The Raskin Center for Humane Interfaces, which he founded after leaving Apple, built Archy — an operating environment designed around the principles of The Humane Interface.

Archy replaced the conventional file-and-folder desktop metaphor with a continuous, searchable document space. It replaced menus and dialog boxes with LEAP, a keyboard-driven navigation system. It was designed to be entirely modeless. It was experimental, never mainstream, but it demonstrated that the principles were implementable.

• Read about Archy →
• RCHI introduction to Archy →

Influence

The Humane Interface is assigned in HCI courses at universities worldwide. Its arguments about modes, habituation, and cognitive load have been absorbed into mainstream thinking about interface design — often without attribution, which is the fate of ideas that become foundational.

The book’s influence is visible in the design of mobile operating systems, which eliminated many of the modal behaviors that desktop systems had normalized. It is visible in the movement toward “flat” interface design, which reduces the visual complexity that competes with users’ attention. Raskin did not cause these developments singlehandedly, but he articulated the principles that underlie them clearly and early.

The Book

The Humane Interface is published by Addison-Wesley. Details:

• Buy a copy →
• Holland’s review of The Humane Interface →

This page is part of the Jef Raskin Archive, preserving the works of the creator of the Macintosh project at Apple and the author of The Humane Interface.