Why we mentor (and should mentor more)
In the Odyssey, the ancient Greek epic poem of adventure and personal insight, we meet Mentor, a wise old friend of the family whose name means “mental strength”, and who gives crucial guidance to Telemachus, the son of Odysseus. (In fact, the Mentor who appears to Telemachus is Athena—the goddess of knowledge, courage, and skill—in disguise.) We all need mentors to help us with complex endeavors, where one wrong move might compromise the entire effort. Classroom learning is a start, but every expert physician, lawyer, engineer, and architect requires one-on-one guidance along the way to develop intuition and precision in addressing open-ended problems not found in books. This is, of course, also true of system dynamics.
In Appendix O of Industrial Dynamics (1961), Jay Forrester spends eight pages describing 25 “Beginner’s Difficulties”; these include overly narrow scope, wrong or vague question, lack of realism, inconsistency of unit measures, excess or insufficient detail, non-robust formulations, and failure to include important delays and nonlinearities. Many articles and books have been written in the decades since Industrial Dynamics with extensive tips for proper model building and testing; and classroom teaching of SD has become more effective. But SD’s pitfalls are many and treacherous, and beginners continue to have difficulties in practice that can only be overcome with the guidance of an expert.
Where can such guidance be found? Some find it from a professor while working on a thesis, or from a senior colleague on the job. But many others do not receive needed guidance—do not serve an adequate apprenticeship. They inevitably fall prey to error, and may continue to do so until some bad result comes to light. When bad modeling comes to light, it not only hurts the modeler, but can also tarnish the reputation of our field generally. This is not just theoretical: I have heard directly about such cases at the highest level of industry.
Clearly, we have a mentorship gap that needs to be filled. Some mentoring occurs at our annual conference in the form of the Modeling Assistance Workshop (which, by the way, has been offered every year since 2001), the Publishing Assistance Workshop, the System Dynamics Colloquium, and the Work-in-Progress Discussions. These activities have helped many modelers over the years, but they are only brief encounters. More continuous mentoring relationships are needed. Perhaps the Society’s Special Interest Groups or Chapters could take charge on this issue, although initial experiments along these lines have been short-lived.
I have heard it said that the supply of willing and able mentors is inadequate to the demand. But maybe this is just an illusion, and we simply haven’t yet found the right way to tap a resource that does exist. For example, many expert practitioners of SD are now getting older and in retirement or semi-retirement, and might welcome the opportunity to be regular mentors, if only they were asked. (They might only require that the mentorship be done remotely, via the web, which is certainly possible.) One might also imagine a “Learning Link” with mentorship requests and offers at the Society website, similar to our successful “Career Link” with job listings.
In the Odyssey, Telemachus needs Mentor as a guide so that he can pull himself together and get past his state of “napios”, or disconnection; disconnection from his ancestors and from his mind and heart. Our hardworking SD teachers and professors start the process, but we also need mentors to help modelers adapt to the nuanced requirements of high-quality SD in practice. Mentors can help beginners connect to what our “ancestors” have set down and apply this knowledge to new situations with rigor and “mental strength.”