Friday, March 4, 2011

Constraint Satisfaction


The concept of constraint satisfaction concerns the constraints that must be taken into consideration when solving a problem or making a decision...the satisfaction occurs when these constraints are met, which leads to a satisfying solution.  Stephen M. Kosslyn, the author of Image and Mind, discusses this concept and provides examples.

The insight that this concept holds is that there are usually just a few ways that a problem may be solved--due to the constraints.  It turns out that by imposing constraints, you can increase creativity in reaching a solution. The reason is that your options become so limited by the constraints that you are forced to seek creative solutions to the problem.  An example might be seen in a TV show on the Food Network...shows in which chefs compete with each other to create the best dish.  The contestants often are given challenges in which they are restricted to certain ingredients (sometimes quite strange ingredients such as mango, octopus, and cornmeal). Similarly, TV shows feature fashion designers who are given challenges to, for example, create a cocktail dress out of plants (leaves, flowers, etc.).

The audience is intrigued because they think, "How can anyone possibly make something that tastes good or is fashionable out of that?" In fact, some of the most delicious or spectacularly beautiful creations come from these challenges in which the contestants are restricted to a few ingredients or materials. I sometimes watch these shows because the process the chefs or fashion designers go through to realize their creations is fascinating. You don't have to be particularly interested in cooking or fashion to learn something about creativity.

There is a lesson here for those of us in science. When our options are greatly restricted, we are forced to get creative, which may lead to an elegant solution. Perhaps we even may discover a new method to address a particular problem.  Those of us who conduct fieldwork in remote locations are often faced with constraints.  We plan ahead, of course, and bring whatever instruments, supplies, and so forth that we anticipate will be required to accomplish our research.  But then, an instrument breaks down, your assistant falls ill, your original experimental design does not address the actual situation, or any other number of constraints. If you are in a remote rain forest, you cannot run to the local Home Depot or Radio Shack to purchase replacement parts. You are forced to improvise, to rethink your research plan, to seek out local guides...to be creative in arriving at a solution to the problem. You think to yourself, "My original plan is unworkable.  What can I do given the constraints of the situation?"

It's quite amazing to be in this situation and to think of a creative solution, which eventually leads to a result that far exceeds what you had originally anticipated. I can think of several studies that I conducted under such constraints...situations in which the original plan had to be completely scrapped.  The end result was an insight that would have otherwise never been realized.  The journal articles that resulted from these situations are some of my most popular (based on citations) and with which I am most satisfied.

I can imagine a new TV show in which scientists are dropped into a remote setting and given a bag of miscellaneous, everyday items and a scientific question to answer.  I realize there have been other shows of this nature, but the participants were typically given unlimited resources to solve the problem.  The difference with my example is that the challenges would occur under very constrained conditions so that contestants would have to be creative.  An interesting twist might be to have a team of Ph.D. scientists competing with a team of non-scientists.  You might think that the scientists would have an unfair advantage, but not necessarily.  Studies show that a diverse group of people (with different educational backgrounds) often out compete a group of Ph.D.s.  The losing team would be forced to vote off one of its members (would it be the weakest link or the strongest?).  It would be interesting to see if the teams evolve to produce groups that are more or less creative, more or less cooperative, and more or less successful in the long-run.  An added benefit would be that the audience would realize that science is not just about high-tech instruments or can only be conducted by highly-trained scientists, but is accessible to anyone who grasps the scientific method and is creative.

Video credit: Moseguaards Cloth Simulation (Alexandra Institute), which involves something called iterative constraint satisfaction (about which I know nothing, but the video was neat).

1 comment:

Anonymous said...

I read something recently about weak links and strong links, and how gender influences which is chosen to go. Men vote off the strongest threat (eliminate their competition), Women vote off the weakest (eliminate the people who are useless for their goals or not team players). The experiments were for mixed-gender and single-gender teams. If I remember where I saw it, I'll post a link.
jc