Abstract of the anti-mac position:
Graphical computer interfaces have become the norm. They are based on a number of principles such as
metaphor, see-and-point, direct manipulation, user control, and WYSIWIG. The Anti-Mac project explored
alternative interfaces that might result from violating the principles behind conventional graphical
interfaces. What emerges is a human-computer interface based on language, a richer representation of
objects, expert users, skilled agents, and shared control.
Keywords:
Computer-human interface, Macintosh human interface, metaphor, direct manipulation, user control,
WYSIWYG, user interface design, WIMP interface, language, computer agents, objects, attributes,
futurism.
Beyond Macintosh
Physicists and mathematicians often stretch their imagination by considering what the world would be like if
some of their basic assumptions and principles were violated (for example, see [1]). This has led to new
concepts such as non-Euclidean geometry, positrons, anti-matter, and anti-gravity. At the least, violating
basic assumptions is a useful mental exercise, but a surprising number of the resulting concepts have ended
up as useful descriptions of the real world.
In the Anti-Mac project, Gentner and Nielsen explored the types of interfaces that could result if we
violated each of the Macintosh human interface design principles. The primary reason for focusing on the
principles behind the Macintosh interface is that the Macintosh is commonly thought to be a good example
of the current interface paradigm and Apple Computer has published an explicit list of Macintosh human
interface design principles [2]. These principles have not significantly changed since the introduction of the
Macintosh; style guides for other popular graphical interfaces, such as Motif, OPEN LOOK, and Windows,
list a very similar set of principles as the basis for their interfaces.
We should state at the outset that we are devoted fans of the Macintosh human interface and frequent users
of Macintosh computers. Our purpose is not to argue that the Macintosh human interface guidelines are bad
principles, but rather to explore alternative approaches to computer interfaces. The Anti-Mac interface is
not intended to be hostile to the Mac, only different. In fact, human interface designers at Apple and
elsewhere have already incorporated some of the Anti-Mac features into the Macintosh desktop and
applications. The Macintosh was designed to be "the computer for the rest of us" and succeeded well
enough that it became "the first personal computer good enough to be criticized" as Alan Kay once said.
The Anti-Mac project should be taken in the same spirit.
The Macintosh human interface design principles
According to Macintosh Human Interface Guidelines [2], the design of human interfaces for Macintosh
system software and applications is based on a number of basic principles of human-computer interaction.
These principles are: metaphors, direct manipulation, see-and-point, consistency, WYSIWYG, user control,
feedback and dialog, forgiveness, perceived stability, aesthetic integrity, and modelessness. These
principles have led to excellent graphical interfaces, but we wished to ask two questions. First, how do these
principles limit the computer-human interface? Second, what types of interfaces would result from violating
these principles?
The Anti-Mac interface
Just as the Macintosh design principles are interrelated and give a resulting coherence to the Macintosh
interface, violation of those principles also points to a coherent interface design that we call the Anti-Mac
interface. The basic principles of the Anti-Mac interface are: a) a richer internal representation of objects, b)
a more expressive interface, c) expert users, d) skilled agents, e) the central role of language, and f) mixed
locus of control. It probably should be admitted here that although we attempted to violate all of the
Macintosh human interface design principles for the Anti-Mac interface, in the case of forgiveness and
aesthetic integrity we had to be content with merely reinterpreting the existing principles.
For a document, a richer representation could include its authors, topic matter, keywords, importance,
whether there are other copies, what other documents it is related to, and so forth. The list of attributes is
similar to those that would be needed by a good secretary who was expected to handle the documents
intelligently. It is not necessary for the secretary to fully understand the document contents, but they must
have a general sense of what it is about and its significance. Richer representation of objects will allow
more intelligent interpretation of user commands, but it will also be reflected in a more expressive interface.
Notice on a bookshelf how books have an endless variety of appearances and yet are all recognizable as
books. This variety adds visual interest and helps us quickly locate a particular book. We should think about
the trade-offs between ease of learning and power in computer-human interfaces. If there were a
compensating return in increased power, it would not be unreasonable to expect a person to spend several
years learning to communicate with computers, just as we now expect children to spend twenty years
mastering their native language. Language lets us refer to things that are not immediately present, reason
about potential actions, and use conditionals and other concepts that are not available with a see-and-point
interface. Another important property of language missing in graphical interfaces is the ability to
encapsulate complex groups of objects or actions and refer to them with a single name. Finally, natural
languages can cope with ambiguity and fuzzy categories.
Conclusions
The following table compares some of the characteristics of the original Macintosh and Anti-Mac user
interfaces:
Bruce Tognazzini's "Starfire" film [3] was a recent attempt to show how a high-end workstation might look
in the year 2004. The interface visualized in the film has several similarities to our Anti-Mac design.
Despite its Anti-Mac aspects, Starfire is still a very recognizable computer with many similarities to current
user interfaces and research systems. Indeed, it should be since it will not be realistic to assume anything
else to be able to reach the market in only ten years. Conversely, we can even argue that there are some
Anti-Mac features in evidence in current commercial products. For example, products like On Location and
SearchIt show autonomy in taking the initiative to index the user's file system when needed and allow users
to retrieve email objects by content and other rich attributes even though each email message may be part of
a large text file when viewed through the standard system interface.
To realize the full benefits from the Anti-Mac approach we would argue that it will not be sufficient to
retrofit its features one at a time to systems that basically follow current user interface architectures. We
believe that an integrated design that builds a new user interface from the bottom up will be more coherent
and will have a better potential for increasing user productivity by at least an order of magnitude. A full
Anti-Mac system will likely have to be based on deep object structures in an architecture that supports
network-distributed objects and detailed attributes that are sufficiently standardized to be shared among
multiple functionality modules. Realistically speaking, such a complete redesign will take some time to
appear. Even if the full system were not to appear for several years to come, it is necessary for detailed
research to begin now to develop the needed features and to collect usability data as to how the Anti-Mac
characteristics should be shaped to truly meet users' needs.
References
1. Alfven, H. Worlds-Antiworlds. W. H. Freeman, San Francisco, 1966.
2. Apple Computer. Macintosh Human Interface Guidelines. Addison-Wesley, Reading,
MA, 1992.
3. Tognazzini, B. "The ‘Starfire' video prototype project: A case history." In Proceedings
ACM CHI'94 Conference Human Factors in Computing Systems (Boston, MA, 24-28 April 1994), 99-
105.
