principles for design and formulae for creativity

The majority of my students were on a masters course on distributed interactive systems. This posed an interesting problem. The sort of systems that are relevant for them are not the traditional 'one man and his screen' systems, but web interfaces, mobile devices, ubiquitous computing. I recall this time last year, writing the April column, and wondering what we should be teaching our students in the 21st Century, precisely because of these radical changes in the underlying technology and devices.

The answer must lie in emphasising general principles whilst demonstrating specific applications. However more and more I find myself describing things not just in terms of general HCI principles, but in terms of general design principles. I continually return to one of the key principles of design - understand your materials. In HCI these materials include the people themselves as well as the devices they use. Again, the focus on purpose and context in HCI are not HCI issues only, but relevant to any design.

Another group of students are doing a research-based masters course, and their module assignment includes an analytic literature review of an area they chose in HCI. Again I keep finding that individual tutorials focus less on the specific topics they have chosen, and more on general techniques to externalise their own reactions to the papers they've read, discovering themes, issues and links within the body of collected material.

Neither general design principles nor these techniques for analytic thinking offer specific solutions to problems or give prescriptions for good system design. Instead they enable the student to be innovative and creative. Although I can't guarantee inspiration I can be 95% confident that they will discover something exciting and new for themselves, but based solidly on established work.

When I was a child I used to look for formulaic answers, mechanisms, themselves created by some inspiration or invention, but which, once so created, automatically and repeatedly give correct solutions to problems. This is perhaps the way of the mathematician, finding, via some mysterious process, timeless truths, that once discovered no longer require insight either of their creator or reader. Mathematics textbooks also take this form presenting theorems and proofs as finished, perfect, items, but not the process through mistaken paths, hunches and insight that lead to them,

Now perhaps I look for creativity and formality more evenly spread - methods, techniques, frameworks that structure and guide, not solving a problem, more putting the person using them into a position where they can add critical insight. If we want to produce software, electronic or mechanical hardware or bureaucratic procedures, these must be reduced to the mechanical. However, design is more contingent yet still constrained.

Looking at other disciplines we see different mixes of creativity and formality. Nowhere is this more clear than in creative arts. In Western (or at least 20th century) culture there has been a focus on internal inspiration, the sense that the artist is born with a muse within. In contrast, when reading about oriental art I found that the eastern artist would reproduce pictures of masters (I'm sure not without adding something of their own flair), for perhaps 40 years before presuming to produce something new of their own.

If we look at music we see perhaps a more balanced view of individuality and creativity vs. discipline and structure. There is no denying the sheer creative power of the greatest musicians, yet this is also accompanied by endless practice, skills obtained through diligence and training. You do not expect to put a flute in the hand of a young child and expect it to produce inspired works in the way we do with a paint brush. Teaching HCI should be more like music than either mathematics or (western) fine art.

Returning to basic principles of HCI, John Carroll is in the process of editing a book on theories in HCI This is precisely drawing out some of the enduring theories that underlie the discipline and should complement Andrew Monk and Nigel Gilbert's excellent edited book, "Perspectives on HCI: Diverse Approaches" (Academic Press, 1995).