What’s the essence of blogging?

Josh Marshall writes the following for the 25th anniversary of his political blog, Talking Points Memo, listing what he identifies as his three principles of blogging:

One: You can’t cover political news well unless you’re familiar with and care about the substance. You can’t cover a piece of new health care policy legislation if you don’t have a solid grounding in health care policy. And you can’t do it well if you don’t care about the outcomes. You can’t cover a Hill story or a campaign well if it’s just a game. Your investment doesn’t need to be partisan. But it should be real. You can’t possibly chart the incremental developments in a story without a basic grasp of how those developments connect up to civic or human outcomes. Otherwise you’re inevitably skimming over the surface, trying to write a novel while being indifferent to the human condition.

Two: Good political reporting should be something you’re dying to read. It shouldn’t be a chore. Headlines should punch you right between the eyes. The story should be as captivating and entertaining and fun as the substance is important. If it’s boring that’s because it’s boring, not because it’s too sophisticated for the audience. TPM should be a thinking person’s tabloid, engaged and engaging. We’re not above and never want to see ourselves as above our audience.

Three: Fundamental honesty with readers. That starts with not lying to our audience. But hopefully that’s a given. What it means to me is that the delta between what we know as editors and reporters and what our readers can learn from reading our pages should be as small as possible. We won’t share rumors we can’t confirm or the identities of confidential sources. And the nature of editing means we focus on what is important and substantial, not every quanta of information we came across. But there shouldn’t be a deeper, insider understanding of the story that isn’t there in the story itself. If we have a set of assumptions or commitments we bring to a story, we should be open about those — not because it’s some admission but because it gives readers a deeper understanding of the story. We should also share what we don’t know as much as what we do. That can mean entertaining hypotheticals and giving readers as much visibility as we can into the process and mechanics of reporting. We did these things originally because it’s the way I like to write about and report political news. But the process of doing this — sharing what I didn’t know in advance — invites reader participation and tips, making the audience into a reporting arm of the publication. It also builds trust and loyalty, which is why TPM has a wildly committed, loyal readership, and why we were able to build a membership business, which is why TPM still exists today.

Most of this applies to our blog here, and indeed I think these apply to the blog format in general. Let me go through these from the perspective of our publication, Statistical Modeling, Causal Inference, and Social Science:

1. We care about the substance. Whether it’s Bayesian computation, American politics, the sociology of science, sports, literature, or Jamaican beef patties, we either have a solid grounding in the area or we’re open about what we don’t know. We’re not the world authorities on everything we write about, and we’re not always experts, but we come into each story with some base of knowledge, interest, and personal experience, and we work from there. And we give reasons for our opinions.

2. We try to be readable. At least, we enjoy writing these posts, otherwise we wouldn’t do it! Not all readers agree . . . but that’s their problem! No publication can make everybody happy.

3. We’re open and honest, we don’t lie, and we don’t hold back. As with Marshall, we don’t go around spilling people’s confidential secrets, but, yeah, “there shouldn’t be a deeper, insider understanding of the story that isn’t there in the story itself.” I agree. We’re clear about our assumptions and commitments, and, as with Marshall, “that can mean entertaining hypotheticals and giving readers as much visibility as we can into the process and mechanics of reporting.” And we get lots of reader participation and tips. We don’t have a membership business, though. I guess maybe we could–statistical insight is valuable, no?–but neither I nor my co-bloggers have gone to the trouble to set up such a business, so this blog is just something we do in our free time, which I think is fair enough for me, given that the duties of my job at Columbia are teaching, research, and service, and I think this counts as service.

So there we are. But writing the above made me reflect upon some other characteristic aspects of blogging:

4. Ongoing stories and plot lines. These range from the general (statistical graphs as comparisons, priors as information or assumptions rather than beliefs) to the specific (Shreddergate, himmicanes, Jamaican beef patties, etc.). Part of blogging is the sense that you’re part of a community, including characters (co-bloggers, objects of admiration, foils, and people who write for other blogs) and story arcs, with a shared history and catch phrases. There’s a balance here–you don’t want to be too insular, you want current readers to feel included without new readers feeling excluded–but I think this feeling of community is important. Marshall notes that his community is so strong that he can support the business using a subscription base. We don’t do that here, but we enhance the feeling of community in this space in other ways: by writing posts in response to readers’ tips, stories, and questions; by providing service such as statistics tips, job announcements, and interesting links; and by engaging respectfully with readers in the comments section.

5. Writing style. Blog writing is informal, but it’s not just that. There’s also a respect for our audience: we feel comfortable jumping into the middle of a story, confidence that readers will look up the relevant background as necessary. You can always go to the search box here and type in what you’re looking for. Back when I was writing for the political science blog the Monkey Cage, it was like that–but then when we became part of the Washington Post, we were pushed to write in a more standardized big-media style. Here was the advice I was given at the time, “to try to professionalize the blog”:

Once upon a time blogging was half-personal, half-riffing, half-informed. (Yes, that adds up to 3/2.) And it was iterative; you said something, you mulled it over later, you came back to it, you assumed a following.

But the model has changed with the advent of googling & social media, which brings people in for a single subject, and with the more serious blogs migrating to major newspapers, which calls on them to retain a bit of attitude, but not quite as personal as it used to be. Now each post is a small essay that summarizes, recapitulates, and moves the public conversation somewhat.

There’s nothing wrong with that–it’s what Paul Krugman does, it’s what a lot of newspaper columnists do. I just didn’t find it so pleasant to write that way. I think that would be the sort of writing that I would do to promote academic articles that I’d written. So in retrospect I guess that’s what I should’ve done for the Monkey Cage: maybe I publish a half-dozen papers each year of political science interest, and each time I publish a paper I could blog on it, write an explainer to clarify what we did and discuss its relevance. This is different from what I think of as “blogging” but it can be useful in its own way.

The other thing is it’s not gonna work if I have to go through an editor. Writing directly for an audience is great; writing indirectly through an editor is horrible. That’s one reason why writing books is fun, while writing articles is a pain in the ass. I’ll go through that review process–for an academic article, the peer review gives it credibility with some readers, and for a newspaper or general-interest magazine, publication gives me the opportunity to reach some large group of readers who I’d never otherwise be able to touch–and, to be fair, editorial review almost always improves what I write–but I do think there’s a benefit to being able to produce large amounts of useful material with little effort, as I can do here and when writing books.

6. No need to come to a strong conclusion. In scholarly journals you’re supposed to have some big finding. In popular writing you’re supposed to tell a coherent story. But most of the time we don’t have a big finding, and lots of stories are not so coherent. Details matter! Of course I say that; I’m a statistician. When blogging we can do lists (as here), we can have ambiguous conclusions, all sorts of things. That’s true of Talking Points Memo too: Marshall and his co-bloggers can and do often come to the conclusion that they’re not sure, they make conjectures while expressing uncertainty, and they don’t need to make each person they write about be either a good guy, a bad guy, or any variants of those categories.

That all said, this doesn’t mean I think blogging is perfect or the best of all possible worlds or even the best of this world. There are lots of things about blogging that I don’t like, including: (a) blogs passing the ball back and forth among a small coterie of insiders, in a kind of mimicry of traditional journalism and traditional academia in which the same few experts keep popping up and agreeing with each other, (b) various blogging tropes themselves becoming accepted unthinkingly (for example, I’ve discussed the concept of “steelmanning”), (c) an aggressive, in-your-face style of writing that I associate with internet gurus, not just bloggers but bloggers can be part of that, and (d) blogs trying to make a case by arguing against something you’ve never said. This last bit is not just a problem with blogging–it happens on twitter too–; maybe the difference is that I expect to see this thing on the hit-and-run medium that is Twitter, whereas blogs give the opportunity for comments, responses, and corrections, but it can happen there too.

So, yeah, blogging’s not perfect. It’s one of many ways to communicate. But, as noted by Marshall, it has some virtues.