This is the first in a series of posts about the human side of science, based on interviews with 27 leading cosmologists. The series explores what drew these scientists to the universe, how they think, what drives them, and what shaped their paths.

01: How this series began

Sometime around 2013, I did something I almost never do. I wrote a fan letter.

The recipient was Alan Lightman, physicist, novelist, and author of Einstein’s Dreams, a deceptively simple yet profound book that I had read many years ago. It was imaginative, poetic and insightful in ways that most science (or science adjacent) writing rarely is. Science in his writing is not a body of facts but rather is a deeply human activity, shaped by imagination and aesthetics and the textures of individual minds.

I had a specific reason for writing. Lightman and Roberta Brawer had published Origins: The Lives and Worlds of Modern Cosmologists, a collection of interviews with 27 leading cosmologists, conducted between 1987 and 1989. When I read the book I could hear scientists talk not just about their work but about beauty, about wonder, about the elegance of an equation or the ugliness of a theory they didn’t trust.

I was leading the Deep Play Research Group at Michigan State at the time, a collective of doctoral students and faculty (co-led with Danah Henriksen) focused on creativity, design, and aesthetics in education. We wanted to hear scientists talk about beauty in their own words, unfiltered. Lightman and Brawer had already done the hard work, connecting with the scientists and interviewing them.

So I wrote to him, explained the project, and asked if there was any way to get the original transcripts.

To my surprise, he replied. To my greater surprise, he agreed. The transcripts were on floppy disks, in WordPerfect format. He didn’t just dig them out and mail them. He took it upon himself to get the contents extracted and sent the files to us. It was an extraordinarily generous act from someone who had no particular reason to help a stranger with an unusual request.

underneath the gunk were 27 conversations with some of the most remarkable scientific minds of the twentieth century: Vera Rubin, Roger Penrose, Fred Hoyle, Stephen Hawking, Allan Sandage, Sandra Faber, Alan Guth, Steven Weinberg, and nineteen others. Over 300,000 words of unedited conversation.

Sarah Keenan-Lechel, a doctoral student in the group, took a first pass through the data. We used some of what we found in presentations and in a book chapter on aesthetics in STEM. But the corpus was enormous, and the real problem was cognitive: no researcher or small team could hold all 27 interviews in mind at once with genuine openness to everything they contained. You could code for what you were looking for. You couldn’t easily discover what you weren’t.

Life moved on. I left Michigan State for Arizona State. The transcripts migrated to a Dropbox folder and sat there. This wasn’t anybody’s dissertation, wasn’t a funded project. It was something I had pursued because I was interested in it, and when the conditions weren’t right, it just sat there… for about twelve years.

A month or so ago, I started playing with Claude Cowork. I did a variety of projects with it most prominent of which was analyzing 2.6 million words of transcripts from the Silver Lining for Learning webinar series I co-host. these experiences showed me what these tools could actually do: not replace a researcher’s judgment, but hold an impossibly large body of text in mind and serve as a genuine conversation partner in the interpretive process. And then I thought: what else do I have sitting around that was always too big to analyze properly?

The Lightman transcripts came back to me almost immediately. A quick search through my dropbox and I had all 27 transcripts, sitting there in a folder on my desktop.

What followed was an intensive four-stage analysis, beginning with completely open-ended reading and building through cross-validation, testing, and full-corpus application. I’ll describe the method in the next post. But twelve themes emerged from the data about the human experience of doing science. The wonder that draws people in, and what becomes of it. Beauty as a working tool, not a metaphor. The startling cognitive diversity among scientists who assume everyone thinks the way they do. The compulsion that keeps them going. The mentors, the luck, the outsider’s perspective. Some of these themes I had been looking for but there were other results that surprised me, revealed only when I finally had the tools to let them emerge.

Much of the physics in these interviews has been transformed by the decades of discovery that followed. But the human material has not. The wonder, the aesthetic judgment, the role of accident and mentorship: these are as true of science in 2026 as they were in 1987.

This series is my attempt to share what those 27 windows revealed. Each post focuses on a different dimension of the scientific life, grounded in the voices of the cosmologists themselves. The transcripts waited twelve years. I think they were worth the wait.

Coming next: How we did the analysis of these interviews—The Artifacts in the Machine.