Subject: Scientific American January 2005

Posted by John Gunkler on 12/18/2004
In Reply To:Scientific American January 2005 Posted by Richard Turnock on 12/17/2004

Message:

Richard, having now read the Scientific American article to which you refer, I have just one supportive comment to make about your letter to the editor.

Much of social science research fails to show strong relationships between pairs of variables. This probably reflects the complexity of the phenomena studied by the social sciences. And it probably reflects what we in the SC community have been saying since Jay Forrester wrote the seminal papers in the 1950's -- to wit, that causation is often circular, that cause and effect are often not close in time or space, and that simple "X causes Y" explanations fail to capture the reality of any system complex enough to warrant scientific study.

Here's another example: In the 1970's (I think), many people tried to show a relationship between worker "satisfaction" (job satisfaction, etc.) and measures of productivity or performance, with (at best) mixed results. This should not have been surprising, except for the strong (unsubstantiated, as it turns out) belief that there "ought" to be such a relationship. The reason such a relationship could not be found, I believe, is because it fails to take into account other intervening variables and it completely ignores systems. To get the gist of this, one only has to examine one's personal experiences. Haven't you known people of both the following types?
1. Happy as a clam, just glad to have a job, to come to work, so see his/her friends every day, and to pick up a paycheck. Only willing to work hard enough to avoid threatening his/her job.
2. Unhappy all the time, never satisfied with the status quo, always seeing things that need to be fixed or change. Works like crazy to do a better job all the time and to help other work smarter as well.

These are obvious counterexamples to the supposed satisfaction-productivity relationship, and not a few people like those described above have populated every workplace I have encountered (and I've worked, as a consultant, in hundreds of workplaces.) I once suggested (before my introduction to SD) that something like "sense of commitment to the success of the organization" combined with "satisfaction" might do a better job of predicting performance. That is, "commitment" was the primary driver, but within categories or levels of "commitment" the higher the satisfaction, the higher the performance would be. I've toyed with a model that captures the dynamics of this hypothesis but haven't taken it very far.

I have one cavil with your letter, however. You wrote "Internal self-talk is the valve that controls how much and how fast the stock of actual self-esteem changes." it seems to me to be premature to make such a strong statement. Perhaps we might better say, "Internal self-talk is possibly a valve ..."

Thanks for the message to this discussion group.