Analogies in Science

How do you carry out your scientific research? And are you doing it the right way? The typical process of scientific inquiry is to come up with a hypothesis, do experiments to prove or disprove it, replicate the results through a number of different techniques and in more than one system, and voila, publish it! The process is simple but there are a multitude of factors that influence decision-making at every step.

One of the most well-used but probably least recognized technique that influences the way we do science is the use of analogy, or the concept of finding similarities between different things. Analogies can ingrain themselves into our psyche often time subconsciously and affect they way we think about science. Laura Otis, American historian in science and Professor in English at Emory, writes in her book “Membranes” about biologist Ramon Cajal who discovered that the brain was composed of individual cellular units. She noted that the method he used to prove this (the Golgi stain) had been invented years before yet no one had used it to define borders between cells. The brain had always been thought of as a big pile of mush, no one would have imagined it was made up of discrete units. She postulates that it was only during the late 18th century, where the idea of colonialism was well-instilled and the philosophy of individual perception was gaining popularity, that people may have been inspired to conceive life in terms of independent living units.

In modern times, analogies are also persistently being used by scientists without them necessarily being aware of it. Kevin Dunbar, a current Professor in Psychology at the University of Maryland, followed four biochemistry labs in Stanford University during the 1990s to discover how scientists think. He came away with the impression that science is an extremely frustrating endeavor (and didn’t we know that already). Of all experiments conducted, 40-60% of them fail. Or to be more accurate, yielded unexpected results. Unexpected results often garner a lot of attention, and are the topic of most lab discussions.

Dunbar found that analogies were frequently used in these lab meetings. 99 analogies at 16 meetings to be exact. He found that analogies could be grouped into classes depending on the goal. Closer analogies, or analogies within the same organism (HIV to HIV) were used to “fix“ failed experiments, analogies between different organisms (HIV to Ebola) were used to formulate new  hypothesis, and analogies between very distant domains such as HIV to a pearl necklace was used when explaining scientific results to a general audience. He also found graduate students made fewer analogies as compared to post-docs or professors, likely due to their limited experience and hence “collection” of analogies.

Dunbar further raised the importance of group reasoning which he termed distributed reasoning. Discussing ideas amongst a group was necessary to learn different viewpoints and thus create analogies. However he said it was ineffective if the group had members of very similar backgrounds (hence only a singular perspective) and also ineffective if backgrounds were too diverse with very different or competing goals. One of the labs that Dunbar followed was composed of the former, a group of E. coli experts. They failed to analogize outside their domain and instead took the approach of modulating different conditions one at a time till the experiment “succeeded”. This however took a far longer time compared to another lab composed of chemists, biologists and medical students which provided more analogies and rapidly came up with a possible explanation and solution.

Interestingly when Dunbar confronted scientists on how they solved the problem, few recognized that they even used analogies! They had little memory of it and instead thought it was the result of their own reasoning that led to the solution. This is probably also why when famous scientists are asked what factors led to their success, many said it was pure luck. It is not pure luck people, it is the use of analogies. Dunbar leaves us with four pieces of advice: 1) Follow up on surprising results, but pay attention to your controls; 2) Use analogies to formulate hypothesis, use distant analogies when explaining things; 3) Go to lab meetings and take part in distributive reasoning; 4) Pay attention to your goals, avoid “confirmation bias” by ensuring your goal is not blinding you to alternate hypothesis.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s