In previous articles, I have often written about "spatial orientation" and "the Spatial Finder." These are not ideas of my own invention, and I have frequently cited other sources, such as Bruce Tognazzini (the founder of the Apple Human Interface Group), as a means to explain myself. In the process, I think my articles have helped to raise the profile of this topic in the Mac community. Today, "the Spatial Finder" is a well-worn subject in many "Mac enthusiast" (and even some "PC enthusiast") discussion forums.
No doubt some would say it is better described it as a "dead horse." Not surprisingly, I object to this characterization. First, in these discussions, there is rarely any consensus as to what, exactly, "the Spatial Finder" is. Many debates (ostensibly) on the topic consist of individuals presenting seemingly opposing points of view based on widely differing premises. Second, intransigence can't (or shouldn't, anyway) keep a good idea down. I still firmly believe that spatially oriented interfaces are too valuable to abandon entirely.
But let's back up and start by nailing down that pesky terminology. I will explain what I mean when I use these terms (which may or may not match up exactly with others' definitions, even Tog's). These definitions will be the basis of later discussion in this article (if not, perhaps, in the wider web world).
"Spatial orientation" is quite a mouthful. It is often easier to talk about the interfaces which take advantage of spatial orientation. One interacts with such an interface by recognizing objects based on their size, shape, position, color, and other visual traits, and then using motor skills to manipulate those objects.
These may be real objects, such as the steering wheel or gear shift in a car, or "virtual" objects such as windows or icons on a computer screen. Obviously, we will be talking about the latter, but it is in relation to the former that the advantages of "virtual" spatial interfaces are revealed.
(Note: Although I will often use the term "spatial interface", very few interfaces are entirely spatially oriented in every aspect. "Spatial interface" is a shorthand description of an interface that predominantly utilizes spatial qualities and relationships as a means of identifying and interacting with objects.)
Let me start by addressing some common misconceptions.
-
Given the above definitions, the conclusion is often drawn that spatial interfaces represent a slavish imitation to real-world objects, and suffer from all the limitations this implies. We've seen many examples of interfaces that attempt to exactly imitate the real world, and the results have not been good.
-
The complexity of personal computers, and file management in particular, has increased tremendously in recent years with the explosive growth of CPU speed and disk storage capacity. Consequently, it has been widely speculated that spatial interfaces, while they have performed admirably in the past, are no longer up to the task of dealing with the volume and diversity of information managed by the modern personal computer.
As I will soon explain, these condemnations of spatial interfaces represent a misinterpretation of the goals and abilities of spatial orientation, and a misunderstanding of the problem domain.
The advantages of spatial interfaces are so obvious that we take them for granted in the real world. From the time we are born, we are presented with the task of recognizing and manipulating objects in space. Even before the time of our earliest memories, dealing with such "interfaces" has become so natural that we never even think of them as interfaces at all.
Knobs and handles on doors, light switches around the house, place settings at the dinner table, even humble pieces of furniture: using all of these real-world spatial interfaces is second nature to us. They all perform well in the following aspects of the broad term "usability."
Ease of learning: these everyday interfaces are easy to learn, partially because they must adhere to physical laws that we all familiarize ourselves with as we grow from infants into children. Physical devices all tend to behave in predictable ways.
Memorability: door knobs and light switches do not move or change on their own, making their location and operation easy to remember. For example, after living in a new house for a few months, the locations of all the commonly used light switches have usually been "memorized." I put that word in quotation marks because this process of memory formation happens naturally over time, mostly without conscious thought.
Efficiency: conscious thought is also rarely involved when actually using these everyday spatial interfaces. We simply enter and leave rooms and buildings; the doors open before us, and the lights turn on and off as necessary. Although we physically perform these tasks ourselves, we do not have to dedicate any real brain power to them. Similarly, eating is rarely a process of conscious "tool use."
Satisfaction: an often overlooked aspect of usability, basic satisfaction with the daily operation of these interfaces is a good indication of their success. With everyday interfaces, "satisfaction" most often presents itself as a simple "lack of annoyances." If satisfaction drops, it is a sure sign that some part of the interface is lacking. For example, we've all encountered "push" doors with handles sticking out of them. Despite the large block letters on the door spelling out the word "PUSH", we often grab the handle and pull long before any thought is given to the process of passing through the doorway. This, needless to say, greatly decreases satisfaction--especially if a lot of people are watching.
It is therefore an admirable goal to imitate the favorable aspects of these real-world spatial interfaces. But a spatial interface does not seek to reproduce all the faults and constraints of real-world objects. Rather, it seeks to exhibit their advantages while transcending their limitations.
But what about the argument that these kinds of interfaces are too simple to be of any use when confronted with the information overload of the digital age? Surely light switches, door knobs, chairs, knives, and spoons exhibit such good usability partially because their tasks are so simple. Much more capable and complex "virtual" interfaces are needed in the computer world, or so the argument goes.
But this argument fails to accurately address the problem at hand: human interaction with large amounts of complex information. The key feature of this problem is not the complexity of the thing to be managed, but the constancy of the thing doing the managing. Put simply, the invariant in this equation is the human being.
The innate skills that humans exhibit when dealing with spatial interfaces have evolved over millions of years out of necessity. Huge (and seemingly disproportionate, given a "common sense" view of "intelligence") portions of the human brain are devoted to processing and interpreting visual information, and coordinating physical actions based on that information.
Perhaps millions of years in the future, the need to deal with physical objects in space will be lessened (or perhaps not--it all depends on which sci-fi author you read). But that time scale is far out of league with the incredibly rapid, technology-driven growth of information complexity. Any human evolutionary progress must be discarded as insignificant when designing interfaces for today.
Moreover, it is a mistake to believe that interface skills evolve on a generational rather than an evolutionary time scale. It's tempting to look at your parents' and grandparents' abilities with technology and conclude that the current and future generations of humans are innately better able to use new kinds of interfaces. But while many things can be learned (vocabulary, functional expectations, traditions, etc.), no amount of study is going to fundamentally reorganize the structure of the human brain in the span of a few generations.
The human mind/body interface is what it is. We must create interfaces that play to its strengths, and avoid its weaknesses.
