What is a business? How can (or should) an expert business practitioner relay his or her knowledge to another interested party? Trying to understand these questions leads down a path of debating the merits and demerits of understanding through metaphors, and understanding through models. This eventually ends up with a discussion of roots in philosophy of science.
During the Seiad project in 1977, Ian Simmonds and I had many discussions about understanding business, filling up the whiteboard in his office at the Watson Research Center.1 My studies into business strategy reflected the two primary foundations: microeconomics — Michael Porter is a leading proponent of this approach — and organization theory — with roots in the research of the Tavistock Institute, and the sociotechnical systems thinking from Fred Emery and Eric Trist. Add onto that my personal bent towards decision support systems — Peter Keen‘s research while at CISR at the Sloan School at MIT was highly influential — and a strategic view of marketing that can be described as Market-Driven Strategy, as described by George Day. These all represent models of business.
Ian — as I recall, starting from a side discussion with Doug McDavid — brought up an alternative approach to businesses, with the book: Images of Organization, by Gareth Morgan. I had a visceral response to this work, because it prescribed the use of metaphors to describe business. The problem that I’ve found with metaphors is that they relay an initial — and possibly superficial — portrayal of business. The layman gets an initial comprehension about the subject, but then starts going off the rails as the metaphor becomes overused. As an example, Stafford Beer wrote books on the Brain of the Firm, and The Heart of Enterprise, that build off models from biological systems. For a information systems analyst who doesn’t understand that a business is a social system — as would be described by Russell Ackoff — what assumptions would be built into the computer system that he or she was instrumental in developing?
The essential challenge, as Ian and I dug deeper forward, was: how to we develop a description of business — in our case, it was an industry reference model for consumer goods distribution and retailing. Doug McDavid had created a more general model in his article on "A standard for business architecture description" in IBM Systems Journal in 1999. Could we get industry people to buy into a more specific version of this type of modeling?
In real life, despite our information-systems-based interests, business people don’t really care for these types of models. Sure, they have their own mental models — as Peter Senge describes — but they’re not really interested in hearing a novice regurgitating rigourous depictions of them. Analysts working towards information systems deconstruct meaning, often drawing lots of charts with bubbles and arrows between them. These aren’t how the experts understand their worlds. It’s how someone who doesn’t understand their worlds tries to express a second-hand understanding.
The interesting statistic — it’s written up by Gerry Zaltman in How Customers Think — is that humans speak at a rate of just under 6 metaphors per minute.2 This statistic has at least two implications:
- charts with bubbles and arrows between them are unlikely to capture the richness of the understanding of content; and
- the transmission of knowledge from a practitioner deep in the domain to another practitioner in the same domain (at the same or a shallower depth) isn’t probably going to be the same as to an "objective observer" who doesn’t share the same practices (and pool of metaphors).
Thus, no analyst — save for one who may also be cross-trained as a practitioner in the domain — is ever going to be able to replicate the knowledge of an expert. The analyst may get close to an understanding … but then what about the views and opinions of second (or third) expert who doesn’t quite see things the same way?
I guess that’s why analysis for information systems becomes an exercise in abstraction. Alternative paths to understand what practitioners really do include Pierre Bourdieu‘s theory of practice, and disclosing new worlds as described by Charles Spinosa, Fernando Flores and Hubert Dreyfus.
1 The depth of the discussions reflected our backgrounds. I spent 8 years in university in the formal study of business: an undergraduate degree in commerce at the University of Toronto, a master’s degree from the Kellogg School of Management at Northwestern University, and two years in the doctoral program in the Faculty of Commerce and Business Administration at the University of British Columbia, before I joined IBM Canada in the headquarters finance and planning function.. In my studies, I had minored in computer science and had concentrations on Management Information Systems throughout that time. My childhood included a rich training in business in my father’s furniture and appliance store in Gravenhurst (northern Ontario), and my career path in IBM had led me through the retail and distribution solutions units in IBM. Ian had studied math at the University of Cambridge, worked in France on the Esprit project, spent some time in IBM’s Toronto Lab, and was settling into the Watson Research Center, on projects related to Insurance Application Architecture. He had done with research with Haim Kilov, with a particular focus on what has become known as Kilov-Ross Information Modeling. In some respects, the challenge could have been described as as a typical "system analysis" question. In reality, because the Seiad project involved IBM Research, we were motivated to develop a deeper understanding of business and information systems, rather than a rushing to a quick-and-dirty answer.
2 I first heard the statistic of "5.8 metaphors per minute" from Jay Ogilvy at a talk at the IBM Advanced Business Institute, on October 14, 1998. In How Customers Think, Gerry Zaltman cites the statistic with a footnote to Raymond W. Gibbs, Jr, "Categorization and Metaphor Understanding", Psychological Review 99, no. 3, (1992).