I write like I'm writing a speech: I imagine myself giving it as a lecture. That's why my AH answers are always filled with crazy amounts of parentheses and especially italics--it's me trying to capture the rhythms of how I would talk it.

In general, I just try to write really clearly and use good topic sentences for each paragraph. Seems to go over pretty well. I mostly get criticized for overusing italics. :)

My advisor, meanwhile, tends to adopt the writing style of whatever primary source he's working with at a given time, in its original language. This is especially hilarious when he's working with something in classical or humanist Latin. The ablative absolutes, they abound.

The first step is to correctly identify the audience you are writing for. You are probably not writing for a "general audience" — there is no such thing, really (there are a multitude of different, fractured audiences), and you probably don't want to write for that broad a public anyway (writing for people without at least a high school education is very hard, as is, if you have a PhD, writing for people without a college education, because grad school writing is generally about writing for people with PhDs).

My term for the audience I usually target is the "NPR audience," which means "educated (largely American, middle class) professionals." It includes the readers of the New York Times, Atlantic, New Yorker, etc. It also includes many academics, which is a bonus, but it is not exclusive to academics. (Consider the difference between this audience and the audience that reads the New York Review of Books — the latter is generally much older and much more highly educated, and as a result a much smaller audience.)

Having done that, you have many models for writing. The key thing is that it is not about "dumbing down" — it is about being clear, avoiding discipline-specific jargon (unless you are willing to devote space to explaining it), and keeping things relatively short (if it is an article, it needs to be under 2,000 words unless you are really doing something quite long-form and special — even then, you need to be really good at your craft if you are going to be holding someone's attention that long, so you shouldn't start by doing that). Books of course are longer but you really need to keep the chapters shorter than is standard in academic monographs.

Because space is so limited you need to make real decisions about whether you want to talk about other scholarship (in general you do less of this unless you really think it is important or that the audience in question will know or care about your "intervention"/"engagement" with exclusively academic concerns), whether you want to bother explaining jargon (or just write so you don't need it), or how many little asides and witty points you want to make. I have found my writing works best when I figure out, from the beginning, exactly what one or two points I want the reader to come away with, and tailor the entire piece around that point.

You also have to get very good at working with editors. A good editor will hack your piece up, remove huge parts that you think are important, rephrase a lot of things (sometimes incorrectly, and you have to be able to find a compromise), and generally make it much better. But academics (in my experience) generally hate strong editing and are not used to it and don't like being told they're a bad writer, that they're unclear, that their argument isn't compelling, etc. Hell, we don't even like genuine fact-checking (how dare you question my knowledge and expertise!), but some places require you to do it. All of this is in my experience very good for the piece of writing but very different than the typical academic writing experience (in which "editing" means a tidying up of the language and a request for a few extra footnote nodding to the right literature).

And, of course, you must read a lot of prose by the kinds of authors you want to emulate and you must pay attention to how they do what they do stylistically. The New Yorker usually has the best historical non-fiction writing of any publication at the moment, in my experience. The works of Jill Lepore and Elizabeth Kolbert are things I pay a lot of attention to. I am a big fan of the book reviews done by Steven Shapin (in the New Yorker and elsewhere). I have gotten much out of the writing of John McPhee though his style is not one I would generally try to adopt (more journalistic). Charles Mann's 1493 is an excellent template for a book-length popular history that both engages with academic work and tells a good story. Eric Schlosser's Command and Control is another recent book that I thought was excellent in terms of balancing the expert and the popular.

Lastly, if you want to write well, you need to write a lot. Writing is a skill, and like all skills it is improved through practice. One does not just sit down and have great writing flow from one's fingers. The more you write, the better you'll get at it. This is one of the many reasons I advocate academics giving blogs a shot — it is a relatively low-stakes place (especially if no one reads it!) to try out different styles and approaches to writing, to developing your own personal "voice." I suspect most people will not develop much of a voice in a vacuum. Writing something relatively compact (i.e., under 2,000 words) on a regular basis (i.e., a couple times a month), with attention to the writing style, will give you good returns after not very long. That sort of regular writing schedule is very different than the standard academic approach.

My article on the Nagasaki bombing on the New Yorker's website is my personal favorite piece of writing that I've done, in part because I really knew what I was trying to accomplish from the beginning and subverted the entire piece to that. My piece on the Trinity test was more of a "throw a lot of things at the wall and see what the editor thinks sticks" sort of piece — I think it does hold together and have a strong theme but that came out of later editing more than the initial work.

This is a topic that is near and dear to me, because I love great literature as well as great history. Unfortunately, in my opinion, only two people managed to write great works of history which were also great works of literature: Edward Gibbon and Winston Churchill. In the modern era I agree with you about Peter Brown: he is a superb writer as well as a superb historian.

I have two different methods I use in an attempt to improve my writing. First, I also write fiction. In fact, I try to alternate between fiction and nonfiction as I publish. I don’t think it is necessary to write fiction for publication, though: they could just be short stories for your own enjoyment.

The rules of fiction are different and allow more flexibility in the text. When you are used to it you can employ some of those same methods in nonfiction text. Even the structure of a history book can be influenced by the traditional plotting you see in novels, and your description of historical individuals can be improved by using some of the characterisation methods of novels.

My second method is that my main editor is actually a fiction editor. He has had a long career in fiction and really knows his stuff. His help has greatly improved my nonfiction prose as well as the structure of my nonfiction books. Under my main editor we use a nonfiction subeditor who is familiar with my field and is able to spot factual errors or problems with my arguments – the main thing historians seem to worry about in regards to which editors they choose.

The most entertaining writers tend to write for humerus but intelligent publications like the New Yorker, The Times, and so forth, so I read those and try to draw some inspiration from it. I have noticed that many historians read the New York Times in particular.

My normal writing style is rather plain and straightforward. Many people say they like my writing because I am able to explain complex subjects simply. I think that is the greatest compliment for an academic: Understanding. If I had to name the author I wish I wrote like the most, it would be Ernest Hemingway. Despite being a giant of literature, nearly every one of his published sentences was a Simple Sentence, and he almost always avoided commas.

T.S. Eliot for phrase, Marc Bloch for punch. If you want an audience interested in sources, write in the style of your sources—Bede (or Felix) for form. And if you choose to pursue an ethos of history (historia—Gk. "inquiry"), pose questions, not answers. A riddle asks more thought than a reason.

I don't know if this is the same, but as a fan and consumer, I think Erik Larson is able to make some really interesting page turners. I haven't looked into on how the majority of his profession thinks of him, but I read The Devil In The White City in 3 days and remember everything about it, bought a book on the history of the world's fair in San Francisco, and sub'd to this sub reddit because it reignited my love of history.

Anthony Beevor sprinkles his books with amazing anecdotes that really drive it home when it comes to the chaos and horror of war. Read any of his books and you will see what I mean.

I think the best prose I've ever read is Thorstein Veblen. He's wry and witty while making brilliant points. I'd give my left arm to write like Veblen.

I really admire Marcus Rediker. The guy is an absolute master at his craft. Villains of All Nations lays out a narrative of pirates as Marxist revolutionaries that is just astonishingly well constructed. We studied it my first semester of grad school and I remain blown away by it. My copy has almost as many post-it flags as pages. He uses Marxian terminology in really sneaky ways, like describing the former slaves as the vanguard, and he builds an argument about the position of women so solid that you might forget there's scanty evidence that a couple of his subjects even existed at all.

I recommend Rediker almost as much as I encourage people to read Hobsbawm.