By Michelle Harasimowicz
An exploration of how linguistic practices impact scientific discovery.
Natural language is the inescapable medium in which we learn and relate information. It infuses our daily tasks: we narrate while searching for keys, think and talk on our feet during meetings, and read to ourselves while glancing at an article. Yet, linguistics is often overlooked in science, eclipsed by the rightfully dominant role of mathematics.
Nevertheless, I would like to emphasize practices for scientific discovery, not rooted in mathematics, but instead in natural language.
I’ll begin by sharing a time when I was studying calorimetry, or the process of measuring the amount of heat released or absorbed during a chemical reaction. I reflected aloud on alternative ways to measure a food’s energy content besides calorimetry, which involves burning the food and measuring the temperature change of a water bath in thermal contact. I mused about different ways that food energy could be described, perhaps by summing up the vast amounts of molecular bond energies, or by applying a voltage across the food and observing how much heat is released by resistance.
I retracted the latter—energy would be added to the system, so capturing the heat from resistance would not be a good indicator of the system’s own energy. Besides that, the food’s resistivity made me realize the lack of uniformity across many foods. It appeared that burning food was a near perfect approach to characterizing the total energy of a non-uniform organic mass, along with its simplicity and succinct thermodynamic equations to back it up.
Moments like this made me wonder why I thought this way. It wasn’t a mathematical line of reasoning—rather, I was attempting to overcome this intellectual puzzle by simply talking myself through it. In writing this article, I set out to identify underlying ideas of what I was experiencing and to learn how to wield natural language tendencies that affect scientific thought.
This led me to interview Dr. Maryellen MacDonald and Dr. Gary Lupyan, both psycholinguists researching language and cognition, and professors of psychology at UW-Madison. Dr. MacDonald studies language production and comprehension, as well as verbal working memory. From Dr. Lupyan’s website, questions are posed like “Does language simply allow us to better communicate our thoughts? Or does it fundamentally shape the structure and format of our mental states?”
I approached both professors with the broad question of how linguistic practices and frameworks provide useful perspectives for scientific work. I will focus on three of the practices they shared: naming, metaphor, and talking.
Naming involves identifying an object by its given characteristics. Both naming and awareness of its impact on cognition can improve science, which is rife with abstract mechanisms and characters. How do you name such abstractions, say an electron, when it is not familiar to us in our everyday macroscopic interactions?
If naming is the ability to define, and defining is the culmination of recognition, to have a deeper connection to the electron’s name, then, you must thoroughly engage the characteristics of the electron through experimentation or calculations. The more you understand and can characterize an object, the more you can operate, model, discern novelty, and spot patterns.
Though, this is probably easier said than done.
“There is no requirement that ideas be expressible. It’s interesting that so many ideas are expressible,” says Dr. Lupyan. This sentiment is palpable in science, particularly in fields like quantum physics, where small-scale behavior doesn’t always mate with macroscopically-derived language.
In addition to naming, we also considered other ways we relate the objects we shape.
“Language is metaphorical through and through—our thought is metaphorical,” says Dr. Lupyan. From the casual to the artistic, metaphors convey a fuller picture of what is being stated. So how can we apply them to science?
Carefully. In the essay “SCIENCE: Magic on the Mind Physicists’ Use of Metaphor,” American physicist Alan Lightman points out: “Physicists have a most ambivalent relationship with metaphor. We desperately want an intuitive sense of our subject, but we have also been trained not to trust too much in our intuition.” A natural language metaphor, then, should be selected as carefully as a mathematical model and should express an idea concisely, while not leading astray to notions of something that is not happening.
Metaphor also has a place in discovery, as metaphorical thought can give a preliminary mental model to test with mathematical rigor and experimentation. A famous example of metaphor aiding in scientific discovery is the story of American theoretical physicist Richard Feynman’s wobbling plates. At a Cornell cafeteria, a rowdy student flung a plate. Feynman considered its motions: “I went on to work out equations of wobbles. Then I thought about how electron orbits start to move in relativity. Then there’s the Dirac Equation in electrodynamics.” Aided by his metaphorical thoughts, Feynman went on to develop equations describing the wobbling plates that later contributed to his Nobel prize-winning breakthroughs in quantum electrodynamics.
Tying these insights together is best done while engaging in the last linguistic practice that I learned about: talking. “Producing [talking] is much harder than comprehending,” says Dr. MacDonald. Talking increases fitness, as does writing. Active engagement with the subject matter is key.
It’s likely that Feynman, a famed science communicator, would have agreed, having championed that the best way to learn is to teach.
MacDonald explained further that talking trains the verbal working memory, the immediate memory of processing information. Verbal working memory is also enriched by long term memory. Someone looking to apply these ideas to scientific discovery should allow time for establishing long term memory, while introducing challenges that enhance the verbal working memory.
Reflecting back on calorimetry, I saw that I had been engaging in these practices. My thoughts became metaphorical as I wondered about ways that food energy could be described. Talking through my attempts to characterize energy in different ways confronted my understanding of the system at hand, a form of active engagement. And recalling the forms of energy I knew was possible due to my long term memory. This funneled into modeling and solving the problems I set forth, where I eventually conceded to the simple elegance of calorimetry.