Take Two: Avi Selk on spiders, aggregation and writing fast

EDITOR’S NOTE: Yesterday, we featured contributor Rebecca Boyle’s interview with Washington Post general assignment reporter Avi Selk about his intriguing, elegiac story of the world’s longest living spider.

Today, we broaden that interview as Boyle delves into Selk’s process for writing fast – very fast – on deadline when his primary tools are the internet and the phone, and how he aggregates already published material with a fresh, authentic voice.

Avi Selk is a reporter on the general  assignment desk of the Washington Post. When he’s not chasing breaking news, he scours the web for intriguing stories he can report by phone and internet research, then weaves his interviews and aggregated material into a fresh narrative. In yesterday’s post about his profile of an Australian spider, he described his job this way:

It’s general assignment, so it’s a mix of breaking news, and when there’s no breaking news, just finding interesting stories — sometimes about spiders. In this specific case, it was actually my editor’s suggestion. My personal preference is away from the darker side, just because I used to be a cops reporter, so I’ve written tons of stories about death and mayhem. But the way the desk works, you take what you get. Sometimes there is stuff that I’m more interested in, and think readers will be interested in, too. And sometimes my editor finds one. These are stories other people have written that we look at potentially doing our own take on.

Most of his stories are done in a day, and the topics range from the tragic to the bizarre. We pick up the conversation by asking about his reporting and writing process. It has been edited for length and clarity.

What else are you working on now?

I just filed a story about a tiger. It’s not nearly as heartwarming. There was a prom in Florida where they hired a caged tiger and brought it out. It was an outrage-type story.

This varies just hour by hour and day to day. I rarely have a project that lasts longer than a day, or have any plans that go more than three hours into the future. It’s fun when you can sort of vary it up like this. It’s been really good for me, because I was sort of like a lazy, slow writer before I got on this desk and wrote maybe one story every month. So it’s been sort of good exercise, like running drills or something. Just to write at that fast pace, it forces you to learn to outline fast and do everything fast and efficiently. If and when I one day become a normal reporter again, those skills will really translate.

You mentioned you write super-fast, so I am curious what strategy you use to do this. Do you make outlines, or do you use some kind of story template?

Even for 300-word stories, I have just gotten in the habit of outlining. I do everything on a laptop, but I will have scrap paper next to me, where I jot down bullet-point level outlines. Sometimes it just says ‘nut graf, then chronology.’ I find that very helpful. I also do outlining simultaneous to the reporting and researching. If it’s an aggregated story, I copy/paste the information into a text file. At that point I jot down a little outline for how I think the story will go, and then do a little more reporting and see if that changes the outline at all.

If and when I one day become a normal reporter again, those skills will really translate.

On my computer, I take every quote or note or aggregated fact and put it all into a single text file that follows the outline. I start every story with this gigantic text file of facts pasted from aggregation or raw notes from my phone interviews. It’s thousands of words long. Then I start winnowing that down into a story with my own prose. I have every single note laid out according to the outline, and sometimes I duplicate that file and start deleting chunks out so it gets smaller and smaller. When it gets to a manageable size, I open up a new file and start writing it in my own voice. I write in a new file as to avoid accidentally copy/pasting someone else’s language in and calling it my own. I write really simply for the first pass. Subject-verb -object. I don’t worry about making it fancy. If that’s all I have time to do, sometimes that’s the way it publishes. I find once it’s all down there in simple prose, I give it another read or two. To what extent I want to have a certain tone or style, that starts coming through on those second passes.

How did you come up with this process?

On this desk you do one story and you finish it at noon, and then do another story. I didn’t have a process before I came to the Post, and sort of figured it out because it keeps me sane and efficient. It just evolved. When I started writing, like everyone else, I started doing it with no plan at all. I could spend half a day staring at three paragraphs.

I need to have my lede down before I can write anything else. Do you have a specific hurdle like that, where you know you need to get past something first?

It’s pretty important to me that I have a good idea of a strong beginning, or how it’s going to flow, or have a strong ending. I have to have that idea, I guess, before I start outlining. But they almost take place simultaneously. If I’m not

When I started writing, like everyone else, I started doing it with no plan at all. I could spend half a day staring at three paragraphs.

entirely sure what my top is going to be, I will write down a couple versions of an outline, and go, ‘Well this doesn’t work, this is going to open in a really boring way. What if I start with this scene instead and loop it back?’ It’s on paper, but it’s really a mental process, too. It helps me think through my lede and my top.

You mentioned aggregating. That can be tricky ground. Do you have sources you know you can count on as reliable? Do you ever worry about tripping up? The Post recently fired another reporter for aggregation that was too close to the source material, so it’s clearly a concern.

I think any journalist has a pretty good sense of what is reliable. To me, a pretty reliable media outlet is one that will run a correction when it screws up. If it doesn’t do that I am really wary of it. It’s sort of a gradient. Any news outlet can screw up.

For me, the biggest one is a story on this march in Poland with a gigantic correction across the top of it. It was this right-wing march in Poland. I was aggregating from CNN and Reuters and AP. CNN reported that there was a banner that said ‘Death to Islam.’ It was a buried detail in the CNN story, but I used it as my lede. Then we wrote a headline around it. Then CNN ran a correction, saying that banner was not actually at that rally. I had the Polish Embassy subtweet me, saying something like, ‘you should be more careful, fake newsman.’ I wrote a first-person essay about how I fucked that up. It’s called ‘Why I Wrote Fake News for the Washington Post.’

One of my editors in Dallas once told me, for any decision just ask yourself: If I fuck this up, and I have to explain how I fucked it up, will it sound reasonable? If not, don’t do it. In that case, I was going off a CNN report. They had someone at the rally. So I don’t mind telling you or anyone how I screwed up, because it happens.

Aggregation used to be verboten, but now it’s not at all verboten. I guess it’s been sort of a messy process to get people to figure out what best practices are. I do a ton of aggregation, but it’s never to be faster. It’s because I can’t get the information otherwise. I treat aggregated material exactly the same I would as raw notes from an interview. When I am

I do a ton of aggregation, but it’s never to be faster. It’s because I can’t get the information otherwise. I treat aggregated material exactly the same I would as raw notes from an interview.

skimming an AP story or something, as much as I can I’m trying to distill the information away from the actual prose. I intentionally sometimes read stories backwards, or just randomly paste lines into a story. If you read what we wrote about Marwa (the reporter dismissed for lax attribution in aggregated stories), she had closely copied not just the information but the form of the story. I would never do that, and I would never want to do that. Mostly I’m too arrogant (laughs). I think I can write stories better than the people I am aggregating. Mostly it’s like, you could get this interview, and I have no way to do that in the next three hours, so I’m going to use that quote.

That spider story had been written in a ton of places, and most of the stories I had read were all parroting each other. It

would have been faster to do it that

way, but I instead read the entire study, and then read the 1970s study that was in the footnote of that study. When you do it that way, and think about your own original outline and how you want to tell it, you end up with something unique.