How I Discovered My Passion for Doing Science

I remember the moment I got hooked on science.

Don’t get me wrong, before that moment I had always liked science. During my K-12 education, I enjoyed my science classes. I found the material interesting. There was a lot I liked about science, as a topic. But it wasn’t until my undergraduate years that I learned what it felt like to do science, and that’s the part I fell in love with.

When I was little, I tried to conduct misguided experiments — such as mixing together substances that I found in the bathroom cabinets — that my father rescued me from. Being that said experiments sprung forth from the whims of a small child, they were dangerous, unprincipled, and did not in any way adhere to the scientific method. In high school, I did “experiments” in my chemistry classes, by which I mean I was assigned to measure reagents, mix them per the assignment’s recipe, observe a reaction, and measure its outcome. While those assignments followed the scientific method in that I was testing a hypothesis derived from a particular theory, and the test was carried out in a principled way, no new knowledge was gained. The thrill of discovery was notably absent. Neither the reckless child nor the jaded high school student were doing the science that sparked passion in me later on.

Enter undergraduate. I started out as a pre-med student, as many college students do, which meant I needed to take lower-level courses in biology, chemistry, and physics. (I never did get around to taking physics.) The first such class I took was cellular and molecular biology, and the lab course that accompanied the lecture. The lab required an independent project working with E. coli bacteria. Specifically, the assignment was to test the ability of an agent of some kind (e.g., a chemical) to inactivate the bacteria. Inactivate, in this context, meant creating an environment that would not support bacterial growth (a bacteriostatic effect). It wasn’t scientific research in the sense that I know it now, but it was an opportunity to come up with a question to test myself.

When brainstorming project ideas with my lab partners, I was struck with inspiration. Reader, are you familiar with Dr. Bronner’s Pure-Castile Liquid Soap? (No, this is not an advertisement for soap. I have no affiliation with Dr. Bronner’s, or any soap manufacturer, for that matter.) The soap is fair trade, made with organic oils, not tested on animals, and even comes in a bottle made from post-consumer recycled plastic. The company supports human rights, the ethical treatment of animals, the fight against climate change, and more. It’s porn to the ethical consumer. All of that is well and good, but the manufacturer also makes some rather bold claims about its uses. Here’s an excerpt from the website:

Dr. Bronner’s 18-in-1 Pure-Castile Soaps are good for just about any cleaning task. Face, body, hair & dishes, laundry, mopping, pets — clean your house and body with no synthetic preservatives, detergents or foaming agents — none! Dilute! Dilute! OK!!

The manufacturer even claims you can use it to brush your teeth! Can you imagine brushing your teeth with liquid soap? I couldn’t. (In the years since my experiment, someone did indeed try to use Dr. Bronner’s magic soap for everything, with surprisingly positive results.)

On top of to the lofty promises of magic soap, the label on the bottle is wild. It is a saga. If you haven’t read it before, I invite you to buckle up and take a ride on Spaceship Earth.

My parents and some friends of the family were sold on the promise of Dr. Bronner’s soap and had used it for years. I, however, was skeptical about a jack-of-all-trades soap that could meet all of your cleaning needs, and was natural to boot. Call me a killjoy. I decided to test whether it could perform its primary job as liquid soap —to destroy germs — as well as a name-brand antibacterial soap (Dial) containing the antibacterial agent triclosan, which has known bacteriostatic effects. Fortunately, my lab partners and instructors were on board with the idea, and we got to work.

The basic experimental approach was to expose a E. coli bacteria in solution to various concentrations of each soap. We used sterile lab techniques when working with the bacteria to avoid cross-contamination from other sources. The primary manipulation was the type of soap the bacteria were exposed to, if any: Dr. Bronner’s pure-castile soap (the peppermint variety), Dial antibacterial soap (containing 0.15% triclosan), or the control condition in which the bacteria were not exposed to soap at all. For each soap, we mixed the soap with bacteria in an Eppendorf tube, then diluted the resulting mixture prior to spreading the solution on an agar plate. In the control condition, we plated the diluted bacteria in solution.

Here are some diagrams I made at the time to help explain the methods. First, we performed a serial dilution of the E. coli bacteria in solution.

A diagram showing a series of 6 test tubes. We pipetted solution between test tubes and added water to dilute further.
A diagram showing a series of 6 test tubes. We pipetted solution between test tubes and added water to dilute further.
No, I was not required to make any diagrams for the assignment, but I was having fun with Adobe Illustrator and, frankly, I’m kind of insufferable.

We performed a similar serial dilution procedure on each soap. In order to pipette the Dial antibacterial soap, which was a viscous gel, we first had to dilute it down to a 10% solution. We then reintroduced the diluted bacteria solution to the soap, and diluted the soap-bacteria mixture even further.

Diagram showing various test tubes that were split into diluted solutions, and a centrifuge that was used at some point.
Diagram showing various test tubes that were split into diluted solutions, and a centrifuge that was used at some point.
Note: Organic soap refers to Dr. Bronner’s pure-castile soap.

Next, we transferred the bacteria-and-soap solution onto agar plates using a micropipette. Agar plates contain a solid growth medium full of delicious and nutritious bacteria food, which serves as an excellent environment for bacterial growth. We used glass beads to spread the solution evenly across the plate, then incubated the plates at 37˚ Celsius for 24 hours. Following incubation, the plates were put in a refrigerator to halt further growth.

Diagram showing a floating hand micropipetting solution from a test tube onto an agar plate, then spreading solutoin around.
Diagram showing a floating hand micropipetting solution from a test tube onto an agar plate, then spreading solutoin around.
Remember when I said we used a sterile technique? In the lab we wore gloves to handle everything, but the art is more fun with ungloved hands, so I exercised some artistic license.

When it came time to check the results, we counted the number of colonies that grew on each plate. We expected there to be fewer bacterial colonies on the plate with Dial solution, since it contained a known antibacterial agent, than on the control plates, which we only applied bacteria in solution to. Since I didn’t believe in magic soap, I expected there to be a comparable number of bacterial colonies on the plates with Dr. Bronner’s soap as the control plates.

To my surprise, that was not the case. The plates we had labeled as treated with Dr. Bronner’s soap had grown zero bacterial colonies. The control plates had grown tens (27.2 on average) and hundreds of colonies (217 on average), depending on the dilution (.000001 or .00001 times the original concentration, respectively), and the Dial plates had about half as many bacterial colonies (103 on average) at an equivalent concentration (.00001x).

Ok, I did not expect that result, but there had to be a simple explanation. Dr. Bronner’s couldn’t have so thoroughly embarrassed Dial like that. We knew the antibacterial soap would kill bacteria (or, more accurately, prevent bacterial growth). It’s in the name, after all! I decided I must have messed up. Clearly, I accidentally switched the soap plates, or I labelled them incorrectly. My bad. Woops! I let my lab group members know what happened and apologized to them profusely. I made sure to be more careful for the second trial. And I was.

We repeated the experiment, taking great care this time to make sure no one made a mistake when labeling the plates. Once again, we performed a serial dilution, this time adding another, lower concentration of bacteria in solution in the control condition.

Diagram shows 7 test tubes used to perform a serial dilution. The first two dilutions are hundredfold, the rest tenfold.
Diagram shows 7 test tubes used to perform a serial dilution. The first two dilutions are hundredfold, the rest tenfold.

We then diluted the soap again and mixed the bacteria solution with it. The solution containing Dr. Bronner’s soap was diluted an additional time, probably because there were still quite a few colonies on the plates that were labeled Dial, but assumed to be Dr. Bronner’s instead. Because I must have made a mistake. Right?

Diagram showing soap dilutions. There’s an eppendorf centrifuge again. We diluted the organic soap a little more.
Diagram showing soap dilutions. There’s an eppendorf centrifuge again. We diluted the organic soap a little more.

We then plated the solutions using glass beads, exactly as before. (I’m not adding a new diagram for it because it would be exactly the same as the last one. You can scroll back up to it, if you require symmetry. I won’t judge.) The plates were again incubated for 24 hours at 37˚ Celsius.

After incubation, I checked the plates and was stunned by what I found. Once again, the plates with bacterial solution that had been exposed to Dr. Bronner’s soap grew no colonies. Zero. Zip. None. I hadn’t mixed up the plates the first time after all! Sure, it’s possible we could have mixed the soaps up a second time, despite being extra vigilant the second time around, but that seemed unlikely. More likely, the soap really was magical. Not in the literal sense, of course, but in the sense that it worked. We were all shocked — me, my lab partners, the instructors — and yet the results spoke for themselves. At the same dilution (0.00001x the original concentration of bacteria in solution), the control plates had the most colonies (394.5 colonies on average), the plates with Dial-treated solution had a handful of colonies (21 on average), and the solution treated with Dr. Bronner’s soap had no colonies at all.

And that, reader, is the moment I got hooked. Not only did I get to test a question I came up with, but I obtained unexpected results, and those unexpected results replicated!

To say it was mind-blowing would be a bit of an understatement. That moment changed the way I thought about science as an enterprise. I shifted from thinking of science as a fact-finding process to realizing science is a creative process that requires insight and ingenuity. As much as I enjoyed science as I had thought of it previously — I mean, those facts were super interesting, and still are! — I enjoyed coming up with new research questions, figuring out how to test them, and seeing what happens so much more. What a rush!

I told my family about the experiment, and even though they had been using Dr. Bronner’s soap for a long time already, even they were shocked by its efficacy. To this day, my dad swears by Dr. Bronner’s soap. And when the topic of surprising experimental results comes up in his teaching, he uses my soap experiment as an example.

But wait, that wasn’t the end of the experiment!

After the second trial, my instructors were a bit concerned because they had used antibacterial soap for the same project in a previous year, and it did a better job preventing bacterial growth before. So, we tested higher concentrations of Dial soap, and we also tried treating the agar plate with Dial directly, then applying the bacteria in solution to the plate afterwards. The higher concentration of Dial (10%, with only one subsequent dilution) prevented any bacterial colonies from growing, which is what we had expected all along. Lower concentrations were less effective (0.1% Dial was about the same as the control).

We also tested Dr. Bronner’s soap on higher concentrations of E. coli bacteria solution, including the full strength solution — containing 100–300 million cfu/mL (cfu = colony-forming units) — and still no colonies grew. Magic.

Here Come the Caveats

As cool as the experience was, and as much as I learned from it, the study we conducted was methodologically flawed.

The study had quite a few confounds, some of which I describe below. For the non-scientists out there, a confound is something that changes along with your experimental conditions, but that you didn’t explicitly control. As a result, you can’t be sure whether any observed effect was due to your manipulation, or if it was driven by the confound instead. While confounds are sometimes unavoidable, a good experimental design is controlled to minimize the presence of confounds or, preferably, to avoid them outright.

The first confound in our study is that, for reasons I can no longer recall (it was over a decade ago!), we tested different concentrations of E. coli in solution across conditions, which made it difficult to compare one condition to another. For the first two trials, there was only one concentration (.00001x dilution) that was used in all three conditions, which means we could only use that concentration to draw comparisons. A better experimental design would have tested equivalent concentrations in all three conditions.

The second confound is that we had to dilute the Dial soap in order to work with it using a micropipette. As our third trial in the experiment showed, the higher concentration of Dial (10% Dial solution) prevented any bacterial growth. The active ingredient, triclosan, is known to work, though it is a controversial ingredient for other reasons. To improve on the experimental design and avoid the confound, we could have diluted the Dr. Bronner’s soap equivalently — the manufacturer says you can dilute it for everyday use, after all — but we didn’t do that, possibly because the soap produced too many bubbles at lower concentrations.

Lastly, we have no way to determine why Dr. Bronner’s was so effective, because we didn’t isolate and test its individual ingredients — or combinations thereof — for antibacterial effects. And the ingredient list is long. Essentially, the ingredients in Dr. Bronner’s soap present a bit of a confound in the study in that we know the soap worked, but we have no idea why — except, of course, that it isn’t actually magic.

There are probably other confounds, too, and there are certainly other flaws with the experiment. Hell, I might not even be using the correct terminology for what we did here, because I’m referencing notes from when I was a nascent scientist who had not yet learned about rigorous experimental design, and I never learned how to write about microbiology in manuscripts. I didn’t become a biologist, I’m a psycholinguist with a PhD in cognitive psychology. I hung up my micropipette a long time ago.

Sure, the moment I discovered my passion for science didn’t occur in the context of publication-worthy — or even presentation-worthy — research. The study wouldn’t have a snowball’s chance getting through peer review. I certainly couldn’t list it on my CV.

Still, I think back on it fondly, and when I do, I remember the excitement, the uncertainty, and the joy of learning something new — not because I read about it in a textbook, but because I observed it myself. That’s the aspect of doing science that I fell in love with.

Postdoctoral researcher studying language and vision at UC Davis. Opinions expressed are mine, not my employer’s.