The How of the Historian

By me. Here at Medium.

Many people — okay, most people — don’t know what historians do. Not, I hasten to add, that’s there’s any reason that they should know. I don’t know what engineers and radial oncologists do. I assume that engineers design and build dams and bridges, but that’s the beginning, middle, and end of my knowledge about their work. I think radial oncologists read images (x-rays and ultrasounds and so forth) although I’m not sure about that.

Historians, on the other hand, I know something about. I am one.

And the historian in me thinks that perhaps “history” would be more palatable (or, gasp, perhaps even respected) if the “public” — that amorphous mass of others-that-aren’t-ourselves — knew a bit more about how historians do what they do. So I’ll take a stab at it (because god knows my colleagues in the academic world aren’t gonna do it).(*1)

When I’m working on a project (and most of my projects are books) (and I’m pretty much always working on a project), my actual “work” falls into one of two general categories: First I research; then I write about what I’ve learned. The research comes first (because I can’t write until I know something) and is the most time-consuming of the two categories.

In my experience, it’s the “research” part of the equation that causes the most confusion among non-historians. Over and over again, people ask me “So, do you Google to find your facts?”

Answer: No and rarely. (I’ll come back to this later.) Instead, the bulk of my research involves reading “sources” that fall into two general categories: primary and secondary.

Letters written by Ernest Hemingway are primary sources. A book or essay written by another scholar is an example of a secondary source. The author of such a source probably used Hemingway’s letters as primary sources. (Confused yet?)

For example, my second book is a history of Key West, Florida. To write it, I needed to learn about Hemingway, who lived in that city for several years. I didn’t want to write or research a full biography of Hemingway; I did, however, need and want to read what he’d said about Key West.

So I read letters that he and his friends wrote during his and their time in Key West. Then I fleshed out my knowledge of Hemingway’s Key West years by reading biographies of Hemingway (and of many of his friends).

The letters (primary sources) provided Hemingway’s first-hand accounts of Key West. The biographies (secondary sources) provided details of Hemingway’s life (his childhood in Illinois, his years in Paris, etc.)

So it is when conducting research in general, whether the history of beer in America (my last book) or a history of the American way of meat (my forthcoming book). To write those books, I read a wide range of documents — thousands of them. What felt, most of the time, like an endless river of documents.

In the case of my new book, In Meat We Trust: An Unexpected History of Carnivore America, for example, the primary materials included government reports, transcripts of congressional hearings, newspapers (from the 1700s on), books, and meat industry newspapers and magazines.

Secondary materials included books and articles written by historians, geographers, anthropologists, sociologists, and others, all of whom had used other primary materials (such as statistical data) to analyze various aspects of meat consumption, production, and so forth. I exaggerate not a whit when I say that I read thousands of documents, from newspaper articles published in 1725 to Michael Pollan’sOmnivore’s Dilemma. (That latter, by the way, I read as a primary document, not a secondary one.) (NOW are you confused?)

As for that question I’ve been asked fifty bajillion times: “So, do you Google to find your facts?” Not so much. Google is useful primarily as a tool for finding other sources.

Say I’m looking for an issue of a magazine published in 1891 and my local university library doesn’t have that issue (or, cough cough, I’m too lazy that day to leave my desk, go to the library, and sit at a microfilm reader for at least an hour). So I Google to see if that issue is available as a Google Book or in the Hathi Trust collection.

Sometimes I use Google to find specific pieces of information. When I wrote the beer book, for example, I wanted to read the entire text of President Franklin Roosevelt’s first inaugural address. Ten years ago, I would have had to go to the university library to find a copy. Nowadays, that’s the kind of thing I can find online.

But in both cases, I’m looking for something specific. I need a specific fact and I know I can find it in a specific document and, most important, I know that I can find an accurate transcription of that speech online. (Because the big caveat about the digital era is crucial for accurate research: Don’t believe everything you read online.)

Research takes time. As in: months and often years. Seriously. (*2)

First I have to figure out what sources are available, and which of those sources will be useful. That alone can take several months. I also must sift and sort as I work my way through this phase of the project. Do I want to read newspapers first? Should I look at government documents first? Diaries? Letters?

Then I have to find those sources. They aren’t piled in a neat stack somewhere, waiting for me to leaf through them. If I’m lucky, I’ll find much of it in the university library in the town where I live. If what I need isn’t there, often the library is able to borrow it for me.

But much of the material that historians rely on is stored in special collections: at historical societies, for example, or at specialized libraries, or in the “special collections” departments of research libraries. And must of that stuff can’t be borrowed. Think, for example, original copies of letters or diaries that have not been reproduced in any form. If I want to read them, I have to go to the place where they’re stored. For the beer book, I visited a dozen or so different kinds of repositories located in Milwaukee, Chicago, and St. Louis.

A great deal of the material I use is only be available on microfilm or microfiche. I’ve spent the equivalent of several years of my life sitting at microfilm readers loading reels of film, slowly scanning the film, re-winding the reels, etc. (*3)

These days, and hallelujah!, many (although certainly not all) primary sources have been digitized, and even when sources themselves aren’t, indexes that can help me sort through documents are digital. (*4)

That’s been an extraordinary time saver. Even ten years ago, if I wanted to read up on a topic in, say, American Farmer, a serial published in the 19th century, I had to go read the table of contents from each issue of the magazine. Now? I can search by keyword. (Although I’m not sure ifAmerican Farmer has been digitized. But you get the drift.) A timesaver? Ohdeargodinheavenyes.

The downside to digitization, however, is that it’s upped the historian’s ante: Now that so. much. stuff. is available in a way that it was not a decade ago, I fear, as do all historians, that I’ll miss something important. I try to balance that fear with this thought: Even if I miss a source or a fact, I won’t miss the main point or the general thrust. (Still, I’m sure I’m not the only historian who sometimes wakes up in the dark hours thinking “Oh, fuck. What if I’ve missed something!?!?”)

In between reading all those primary documents, whose number is infinite, I’m also reading secondary stuff: Who’s written what that will fill in the gaps in my knowledge and can provide material for background? (See my Hemingway example above.) What have other scholars and experts said about this topic? Does my view jibe with theirs? Am I about to become an intellectual outlier and thus the subject of ridicule? (God forbid.)

The filament that connects all this activity is the brainwork: As I read, I learn, right? So each source or document I read informs the next, and my knowledge about and understanding of my topic is expanding by the moment, each bit and byte of information adding a layer to the whole.

For me, the first year or so of a new project is total confusion. I typically write about topics about which I know nothing, so when I say I’m “learning,” I mean that literally. As we all know, the early moments of learning are ones of bewilderment. “Now what does this mean? How do I do this? What am I supposed to do next?”

So it goes with all these sources. And so goes the process of research: I read. I absorb. I move to the next source. Read. Absorb. Repeat. Hundreds of times. Thousands of times. (I’m not complaining, by the way. I love my work.)

Gradually, and mercifully, bewilderment gives way to understanding and after months, even years, I feel as though I’ve “learned” enough so that I’m ready to start writing.(*5)

And then things get interesting. With a capital I.

As soon as I try to write about what I’ve learned, my brain begins a conversation with itself and with those hundreds/thousands of sources I’ve read. I begin to understand what my book is about and what all those primary and secondary sources mean. And of course the meaning of those sources changes as I begin to examine them in relation to each other.

As my brain filters that information, I respond. In practical terms, that means I must return to my sources and re-read them because I’ve decided that my initial understanding of them was flawed. Or I discover that I need to read different sources, ones that I didn’t realize were relevant.

So off I go to the sources. And then I return to my keyboard — and the process starts all over again.

And that, dear reader, is why it takes me so long to write a book. Five years for the beer book. Seven for the meat book.

Years!” you say. “How can it take you years to write a book?” Look at Michael Pollan or Thomas Friedman or Doris Kearns Goodwin? They publish a new book about once a year. What? Are you stupid? Are you lazy?”

Maybe and maybe. I dunno. I do know, however, that those people have something I don’t have: Help. They pay assistants to do the scut work, much of the research, and, for better or worse, much of their writing, especially of the first drafts. Goodwin and Pollan don’t have to make umpty-bajillion trips to the library. Hell, they don’t even have to scan the indexes and compile the articles and congressional hearings and expert studies. Someone does all that for them. (My weak imagination can’t fathom that kind of help.)

And there you have it: The historian’s life. And now you know: Google isn’t enough.


  • *1: As a historian, my primary goal is to bring good history to, as I put it, “the rest of us.” Among academic historians, people like me are known as “popular historians.” A bit misleading, that bit of phraseology: it doesn’t mean that I or my work are popular. Rather, I write for a general audience rather than a scholarly one. And yes, people like me are regarded with contempt by the “professional” historians. I have the same credential as them — a Ph.D. — but I don’t have an academic post and therefore my work isn’t quite as “real” as that of “real” historians. (If you figure out why, do let me know.)
  • *2: In no way, shape, or form am I complaining, by the way. I LOVE my work.
  • *3: Mercifully, most places these days have machines that wind and re-wind with the push of a button. In graduate school, I actually damaged my right shoulder cranking the handle of a microfilm reader. (I’m a woman of limited imagination. I can’t make up stuff like that.)
  • *4: This is why I fall over in a heap howling with laughter whenever I hear the two Google Guys saying that they want to digitize everything. Do they have a clue? (Answer: no.) Do they have ANY idea how much “information” there is the world?? And how much of it is stored on pieces of paper piled high in archives around the world? Pals, you ain’t gonna be digitizing “the world’s knowledge” any time soon. Not unless you’re prepared to turn over your immense fortunes to the librarians and archivists of the world.
  • *5: How, you may wonder, do I know when I’ve learned “enough” to move on to the writing? Because the stuff I’m reading starts to “read,” feel, and sound familiar. There aren’t quite so many surprises; the number of unknowns decreases. It’s an intuitive thing.