Design through Analogy
“I especially love analogies, my most faithful masters, acquainted with all the secrets of nature… One should make great use of them.”
Johannes Kepler was brilliant. In fact, if there was no Kepler, there weren’t have been a Sir Isaac Newton. Or at least Newton’s discoveries might have come much later and perhaps by someone else. After all, Newton stood upon the shoulders of giants, and one of those giants was Kepler.
What makes Kepler so fascinating is that he figured out gravity before Newton. Not a lot of people have that on their resume. No, Kepler didn’t define all the fine details of universal gravitation, but he was able to solve planetary motion without understanding Newtonian physics. Kepler’s laws of planetary motion were published almost 70 years before Newton’s Philosophiæ Naturalis Principia Mathematica, often referred to as Principia.
Of course, as the scientific discovery chain extends, Einstein later refined Newton’s classical mechanics for extreme masses and speeds, but in the other direction, Copernicus’s heliocentric view was refined later by Kepler himself.
Still, how did Kepler do it with so few tools and with no Calculus? Through analogies.
Kepler loved analogies. He once stated: “I especially love analogies, my most faithful masters, acquainted with all the secrets of nature… One should make great use of them.” [1]
He knew that the heat of a fire decreases as you stand farther away. He knew that smells dissipate the further you retreat from the source. He saw the light of a candle becomes stronger as you approach it. All these analogies helped him start to piece together an “invisible power” somehow linking the planets together.
But he didn’t stop there. He contemplated magnetism as he saw planets appear to be both attracted and repulsed as they traveled closer and farther from the sun. He thought about rivers catching boats in circular currents or eddies. He considered these and many other analogies to help him sort through the limitations and to have theories that match the evidence he was observing. This led him to his amazing discoveries related to planetary motion, but also regarding lunar tides, which, at the time, would seem absurd to declare that the ocean level of a 17th-century harbor was affected by the moon.
Analogies are so important to progress that some could argue it is the only way to creatively solve problems. Albert Einstein is often quoted stating: “We cannot solve our problems with the same thinking we used when we created them.” One way to interpret this quote is that the problems generated in physics can’t be solved by only using known physics, medical challenges won’t be exclusively solved by medical professionals, and social problems won’t be solved by only using tools the politicians and social scientists know, understand, and use. The logic continues that if the tool was already within the domain to solve it, it wouldn’t be an existing problem. Thus, problem solvers necessarily must come from outside the domain, or at least the thinking of the problem solver comes from the outside, that is, from a new domain. Ergo, analogies.
However, analogies go beyond problem-solving. We often need analogies to learn. In engineering, an electric circuit can be taught as a water pumping system with water pressure, water current, and debris in the pipes that restrict the passage of water. The reverse is also possible for the experienced electrician trying to figure out how a sprinkler system operates. Computer programming makes use of extensive analogies including common physical objects or principles, like containers, wrappers, inheritance, and threads. Many statistical or computational processes were likewise bio-inspired. This is why we still use terms like neural networks, genetic algorithms, and swarming for optimizing, classifying, and modeling processes. In some cases, the mathematical adaptation of the physical analogy is nearly equivalent (e.g. ant colony optimization). When seeking help from an instructor to better understand a concept, one of the first strategies to consider is to have them try to explain it with an analogy. Learning and analogies go hand in hand.
Designers are likewise the beneficiaries and applicators of analogies. Every week or so, it seems like another natural phenomenon such as an endangered animal’s attribute, a plant’s features, or a weather event is used in a new design. Of course, this is nothing new and has been going on for a long time. Bird wings clearly inspired aerodynamic designers with their early airfoils shapes, tornadoes evoke ideas for vortex suction in vacuum cleaners, and sharkskin has prompted designers to reduce the drag of bodies through fluids whether that body is a human at the Olympics or a sea-faring vessel. A lot of the key aspects of the usefulness of products originate from a designer temporarily looking at something far removed from their own experience or domain.
None of the foregoing means that we can’t provide incremental progress and contributions in our own domains through many iterations (See Chris’s article on iteration). In fact, a majority of our work may be making many small changes to existing systems or common products. But even these small improvements and iterations are founded on analogies if one thinks about it sufficiently. The key to unlocking a product’s potential (even in small ways) will be in continually applying and using analogies for design.
Essentially, the more expansive one’s access to analogies, the better. An experiment conducted by Karl Duncker, a cognitive psychologist in the first half of the 20th century, supports this finding. He would have individuals try to solve a problem [2]:
Suppose you are a doctor faced with a patient who has a malignant tumor in his stomach. It is impossible to operate on the patient, but unless the tumor is destroyed, the patient will die. There is a kind of ray that can be used to destroy the tumor. If the rays reach the tumor all at once at a sufficiently high intensity, the tumor will be destroyed. Unfortunately, at this intensity, the healthy tissue that the rays pass through on the way to the tumor will also be destroyed. At lower intensities, the rays are harmless to healthy tissue, but they will not affect the tumor either. What type of procedure might be used to destroy the tumor with the rays and at the same time avoid destroying the healthy tissue?
Researchers found that only 10% of participants would come up with the solution of focusing multiple lower-intensity rays from different directions so that they would intersect at the tumor’s location and then collectively have a high enough intensity to destroy the tumor (we engineers would likely call this superposition of signals). However, the researchers found that a higher percentage of participants could arrive at this solution if they were first primed with a story recounting a military general attacking a city with many small armies that all converge on the city at the same time from all directions.
A military officer attacking a city has nothing to do with killing cancer… or does it? Designing by analogy would maintain an emphatic yes.
This is why we should never stop learning and maintain our curiosity about everything (or at least as much as we can) throughout our lives. We never know when a fact, statistic, book, or experience 5, 10, or 20 years ago will come in handy. Someone may make those connections only after a wide and disperse set of experiences and knowledge and therefore would have access to analogies (see Cross Pollinator article). Similarly, this is why closing our minds off to someone’s idea is not just disrespectful in the moment but makes us worse designers. Even if we don’t agree with them, we should still listen and file away their perspective, even if for a selfish reason to later design our own solution to a problem in our future.
References
[1] Quoted in: Epstein, D. J. (2019). Range: Why generalists triumph in a specialized world. New York: Riverhead Books. The discussion in the following paragraphs of Kepler’s strategies to discover planetary motion was also inspired by Chapter 5 of this same book.
[2] Duncker, K. (1945). On problem-solving (L. S. Lees, Trans.). Psychological Monographs, 58(5), i–113.
