It may seem like I’m just playing on words about how the stress of grant writing causes ulcers. That is true, but this message is much deeper than that.
It’s about the story of how great science is done, and what that has to do with your grants.
One of my favorite stories in science is that of Barry Marshall and Robin Warren.
Back in the late 1970s, ulcers were considered a stress disease. Too much acid, too much tension, too much Type-A living. The idea that bacteria could survive in the stomach — let alone cause ulcers — was dismissed outright. It violated the entire operating system of gastroenterology at the time.
But Warren kept seeing the same strange, curved bacteria in stomach biopsies. Not once. Not occasionally. Repeatedly. He saw a pattern that the field wasn’t ready to acknowledge. He imagined that these bacteria were more than just contamination in his samples, and more than just an opportunistic infection. He imagined that they may play a deeper role in the sustenance of the ulcer.
Warren’s observations might have stayed a footnote in pathology if not for one person:
Barry Marshall was then a young internal medicine trainee with more curiosity than credentials.
Marshall entered the picture in 1981. He was assigned to help Warren follow up on these “mysterious curved bacteria.” The more he read, the more he found himself imagining something the field refused even to consider:
What if ulcers weren’t caused by stress at all?
What if the entire model were backwards?
That question — an act of imagination as much as science — hooked him.
Together, Warren and Marshall gathered biopsy samples, cultured what they could, and started seeing consistent patterns of inflammation wherever the bacteria appeared. Every observation strengthened the imagined model taking shape in their minds: microbes → inflammation → ulcers. But every time they tried to present these findings, they hit the same wall:
“Impossible.”
“Contamination.”
“These organisms can’t survive acid.”
“Ulcers are caused by stress and lifestyle. End of discussion.”
Marshall later described the atmosphere as “polite dismissal,” the kind that shuts down a new idea without ever openly engaging it.
They attempted to publish early findings — and reviewers rejected them repeatedly.
They tried to grow the bacteria — and for months, nothing seemed to work because the cultures were routinely discarded after two days, the standard protocol.
But they kept believing in their model. As with all feats of scientific imagination turned into reality, they persisted despite the critiques and their own doubts.
The breakthrough came when a set of plates was accidentally left to incubate over the long Easter weekend. (It sounds reminiscent of Alexander Fleming’s discovery of penicillin — another case where a culture plate left to sit during vacation revealed what no one had imagined.)
When Marshall returned, H. pylori colonies were finally visible!
Still, the broader field didn’t budge.
By 1983–84, Marshall reached a breaking point. His clinical observations were lining up. The pathophysiology made sense. The biopsies were consistent. The cultures were real. But without a definitive human challenge study, the medical establishment wasn’t going to move.
And of course, no ethics board would approve deliberately infecting someone with a potentially disease causing organism based on his model.
The field just couldn’t imagine it. (How many breakthroughs have died because someone couldn’t imagine what was right in front of them?)
Having nothing left to lose, Marshall made a decision that has since become scientific legend.
He drank a beaker of H. pylori culture.
This wasn’t a reckless stunt; it was an act of frustration mixed with conviction that his model -- the one that had started as a tiny seed of imagination -- was pointing to something real.
Marshall believed the only way to break through the resistance was to generate undeniable evidence.
The next part is less well known:
His wife had no idea he was planning this.
When he became violently ill days later — nausea, vomiting, halitosis, stomach pain — she was furious and terrified.
According to Marshall, she insisted he go to the hospital immediately, convinced he was dying. The clinicians who evaluated him noted the rapid onset of gastritis. A biopsy confirmed inflammation. It wasn’t a full-blown ulcer, but it was the precursor.
Marshall then began antibiotics (metronidazole and bismuth, at the time), and the symptoms resolved — providing a complete demonstration of H. pylori–induced disease followed by antibiotic cure.
He wrote up the findings.
He presented them.
Eventually, the field could no longer ignore the evidence.
But acceptance didn’t come quickly. It took:
more clinical trials
more biopsies
more treatment studies
more epidemiological evidence
More senior researchers are gradually admitting the old model didn’t fit the evidence
By the mid-1990s, the consensus had shifted. The field was starting to join Marshall and Warren in the new model of the disease, the one that had started as a small seed of imagination more than a decade before.
Stress and acid were no longer the main villains of ulcers.
H. pylori had moved from “impossible” to “central.”
In 2005, more than twenty years after Warren’s first observations, Marshall and Warren were awarded the Nobel Prize in Physiology or Medicine “for their discovery of the bacterium Helicobacter pylori and its role in gastritis and peptic ulcer disease.”
This Nobel wasn’t about a bacterium. It was about two scientists who trusted their imagination long enough for the rest of the world to finally see what they’d been seeing from the start.
What has this story got to do with funding? If you’ve been going for funding, you may have already had the thought: reviewers never would have funded this work in the first 10-12 years or more. They’re too conservative, especially at funders like the NIH and other major national funders across the globe.
If you had that thought, I agree. Such a major, paradigm-shifting act of imagination like this would not be initially fundable by most funders. It challenged the existing paradigms and scientific conservatism too much.
Yet, just because this wouldn’t likely work directly in a grant proposal doesn’t mean the underlying principle is irrelevant.
In fact, it’s the part most proposals are missing.
Great grants aren’t paradigm revolutions — but they always contain the seed of one.
Not the full Marshall-and-Warren model, but the beginning of it — a coherent, imaginative architecture that stretches the field just enough to feel meaningful, while staying inside the conservative tolerances of the funding mechanism.
Reviewers respond, often without realizing it, when they can feel that you actually see something: a deeper pattern, a clearer model, a direction the field hasn’t articulated yet but instinctively recognizes as true.
You’re showing them how you will move the field forward from where it is, and that always requires a leap of imagination.
It’s not data.
It’s not techniques.
It’s not personnel.
Not at the beginning.
It starts with imagination. Everything else arrives later — the people, the techniques, and the data that justify it. (And if your proposal is about a technical advance, the same principle applies: the technology has to emerge from a conceptual advance if you want reviewers to buy in.)
When that imaginative foundation is missing, a proposal reads like isolated experiments wandering around in search of a story.
Reviewers feel it instantly, even if they can’t articulate why.
And this is the part almost no one teaches: a proposal’s architecture begins in imagination, not in data. It begins by having a clearly defined mental model (the underlying scientific architecture that makes your proposal coherent) that moves the field forward in ways it cares about, supported by methods that can actually test it.
If you skip that step, nothing you build on top of it can compensate.
Most people I’ve worked with reverse that order by default. They start with the methods they want to use or the data they want to collect. They may be carrying around a subconscious, unarticulated model of why those methods and/or data are relevant, but in the failure to articulate it clearly, multiple problems occur:
Once you build the model first — consciously, explicitly — everything else becomes easier, clearer, and more powerful.
Since one email can only sketch the outline, I’m offering a short workshop where we’ll go deeper into this. I’ll walk you through the structure we use with clients to help them formulate clear mental research models — the kind that make proposals more compelling, more coherent, and easier to write.
