How Do We Know What We Know About Coffee—And Life? (Part One)

Many well-worn “truths” about coffee are anything but—and failing to interrogate our sources of knowledge dooms us to repeating old mistakes.

Photo by Krišjānis Kazaks on Unsplash

When I first started working in coffee, I became obsessed with learning the rules: about how coffee should be brewed, about the best way to serve coffee, about how to get the most out of each bean. I wanted desperately to understand the rationale behind what we did every day, to figure out how we knew what we knew.

Making a cup of coffee is like a science experiment, one that’s easily replicable. Even though there are multiple variables at play (from a coffee’s country of origin to water temperature, grind size—heck, even the humidity inside your house), brewing coffee still provides a pretty linear way to establish ideas. I ground my coffee this way, and this happened. The next day, I ground it slightly differently, kept everything else the same, and got a different outcome.

That might be a simplification of how brewing coffee can work, but it fed my assumption that what we know about coffee had been tested, and was the result of some empirical proof. When someone told me a certain brewing technique or processing method was better, I believed that to be a fact.

But over the last 14 years in the industry, I’ve seen those supposed standards change wildly, and the seeming truths we thought we knew about coffee be debunked entirely. I’ve seen folks return to methods once deemed undesirable, and question why we threw away one concept in favor of another. Compared to the certainty of my early coffee career, this flux has made every supposed fact about coffee—everything I thought I knew—feel unstable.

With these thoughts in mind, and using the lens of coffee, I want to explore how we know what we know more broadly. This essay—the first in a two-part series—explores how we establish knowledge by looking at cycles of repetition, how ideas get “rediscovered” and sold as new, and how misinformation becomes fact. In the second essay, which will come out next week, I’ll look at establishing knowledge as an everlasting project, how repetition can build deeper understanding, and the idea that lessons are never just learned once.

Human Touch vs. Automation

I began working as a barista in 2010. At the time, handbrewed methods were an ascending coffee trend. Just a few months before I started my first coffee job, Oliver Strand—a New York Times writer and venerable observer of all things coffee in New York—wrote a piece about how pour-over brewing methods were beginning to infiltrate the best coffee shops in the city. “As coffee-brewing techniques go, pour over is slow and mannered,” he wrote. “It’s low tech. It has a funny name. And yet, pour over is an ongoing obsession within the coffee world. It’s been around for years, though interest has spiked in recent months.”

Pieces like this one made me assume that anything “automated” was lesser in quality—that it took away from the barista’s skills and produced inferior drinks. I worked at a shop where we weren’t allowed to mess with grind settings, and all the shots we pulled were based on pre-set volumetric outputs (you press a button, and a set amount of water runs through the espresso puck and then stops on its own). I understood these settings to be implemented by shops that didn’t have the time (or care) to train their staff. The baristas who worked at places where they constantly tinkered with the coffee seemed somehow more in tune than I was.

I’m sure I wasn’t alone in associating manual tinkering with quality—I recall many of the era’s futile attempts to define “third-wave coffee” placed emphasis on its handmade nature. It seemed irrefutable that pour-over coffees were just better, that having control over when to stop a shot made the espresso superior to anything machine-measured.

Machinery and robotics were key anxieties in the mid-2010s coffee industry. This 2016 article from La Marzocco USA references at least two “experiments” run by folks wanting to compare manual brewing methods to automated systems. There was even an event called “Man vs. Machine” where Nick Cho, after writing an op-ed for Serious Eats expressing his disdain for automatic brewing devices, went head-to-head in—and narrowly lost—a brewing competition with an Alpha Dominche Steampunk machine. (Writing this sentence today feels like writing about dinosaur fossils: Nick’s article is no longer on Serious Eats’ website, Nick is now a social media star, and Steampunk brewers no longer exist; Alpha Dominche ceased operations in 2018. Here’s a video showing how the Steampunk worked).

The article, which was a promotional blog post on the La Marzocco USA website, talked about how the company’s machine, the Linea PB, allowed LA-based G&B Coffee (who, along with employees from Go Get Em Tiger, unionized in December 2023 and whose co-founder, Kyle Glanville, stepped down as CEO in February 2024) to use volumetrics to achieve more consistency and provide faster service. “Volumetrics allow the barista to cover a wider area and do more things and do the central role of pulling shots more accurately,” co-founder Charles Babinski told La Marzocco (Babinski left the company in 2020).

At the time, this felt like backtracking. I had already worked at a shop that operated on programmable volumetric settings, and believed that I was making inferior coffee compared to the baristas with free rein over their machinery. But the tides were starting to shift, and as I progressed in my career, I heard more and more baristas espouse the advantages of batch-brewed coffee. By 2017, I had eliminated pour-over coffee from my small shop in favor of  batch brew. 

Of course, part of the recognition that some automated processes could do better than people resulted from technological advances. But this sea change also made me question why we so vehemently adopted pour-over methods, and believed them to be superior, in the first place.

No Footnote, No Methodology

The above is one small example of how knowledge that feels fixed can ebb and flow. We see this cycle happen all the time: New generations discover what past generations left behind; new technologies or changing context make possible what was once unfeasible. In some ways, this is the natural course of life.

But sometimes, ignoring history can lead to more sinister consequences. Questioning how we know what we know is an idea that arose during the gender and women’s studies course I took about food politics (that class has also inspired several recent essays). 

One of the ideas we discussed was sustainability, and how it might feel like a newer concept since so many brands and companies are increasingly making public proclamations about their eco-friendly and environmentally conscious practices. However, sustainability has been part of Indigenous groups’ practices for centuries, even though that knowledge and know-how is little recognized today.

Failing to ask how we know what we know can also propagate misconceptions about the world around us. In 2019, I watched a talk by Vanusia Nogueira, then the executive director of the Brazilian Specialty Coffee Association and now the executive director of the International Coffee Organization. She talked about how an oft-quoted statistic—that there are 25 million smallholder coffee producers—is actually inaccurate:

“We estimate that we are around 25,000,000 producers in the world. That is an estimate that is shown to us and they just wanted to research that this is not a correct number. The number of the producers nowadays, it’s 50% of that, it’s 12 and a half million producers…”

She based this correction on findings from David Browning of Enveritas, who would give a talk later that day analyzing how this number has been repeated in stories shared by supposedly credible organizations, from the New York Times to PBS. He even points out that the Specialty Coffee Association named its magazine “25” because of that erroneous figure:

“...this number is confirmed by many reputable sources ... It’s all footnoted, but whose estimate is this? And if I click the footnotes, what do I find? If I click that footnote and I trace it back to the original source, what I find is a handful of papers written around 2001 with numbers but no source. A single paper in 1999 with no source and a singular blog post from 1995 offering ‘20 million’ with no source and then the trail runs cold. 

So, to the best of our knowledge, there is no footnote. There is no source of this data, no methodology.”

In his talk, Browning discusses developing a methodology with his team to determine a more accurate number of smallholder coffee producers; they came up with about 12.5 million. He explains how they created the methods and some of the shortcomings of this estimate. And yet, when I did a Google search for how many smallholder coffee farmers there are, I still got results that cite the old number. Some (including Google’s AI assistant, Gemini) cited Browning’s number, but websites as recently as this year still cite the old one.

Browning then goes on to debunk some other commonly accepted “truths,” including the claim that we only use 10% of our brains (we use 100%) and that coffee is the second-most-traded commodity in the world (that’s based on a source from more than 60 years ago, which Browning points out is determined from very shaky data). He argues that he’s not there simply to prove that seemingly established facts are untrue, but rather to point out that without good data, we can’t really begin to solve the problems around us:

“What do we care of things of fact or fiction? Well, it doesn’t matter if you don’t plan to do anything about it, but if you plan to work on the issues, if you’re interested in engaging in problem-solving, it would be far better that we understand what the issues are, what the priorities are, and get to the insight and understanding below the headline. Otherwise, we could spend a lot of time talking and a lot of time working on solutions to problems that aren’t even the most pressing priorities.”

Browning distills the stakes of failing to investigate how we know what we know. Finding a solution fit for 25 million smallholder producers is a very different task than finding a solution for half that number. Without a firm understanding of the world around us—or at least the limitations of the knowledge we think we have—we’re doomed to repeat the mistakes of those before us. When we don’t understand why we’re eschewing a practice or idea, we might find ourselves revisiting it again, unsure why we let go in the first place. What looks like progress might be more like trudging in a circle.

That’s not to say that once an idea is established, it doesn’t need to be learned again. Part of how we know what we know is through repeated experiments and attempts. In next week’s article, I’ll talk about what I believe is critical in establishing knowledge: repetition.