When I think about it for any reasonable amount of time, the concept of interviewing, or of being an interviewer, starts to feel exceptionally weird. At least, I think it’s weird if your goal is some sort of authenticity. To use verbiage that could have been in an
essay, the “technology of the interview” somehow prescribes the sort of interaction you and an interviewee are about to have.You want people to tell you about themselves and their work, and to be somewhat authentic about it, but you all know that whatever words you exchanged are going to be packaged and distributed in a file, etched in digital stone. You’re speaking to one another, but at another level, you’re really not. You’re trying to convey something to the audience who will eventually listen to your words: a message you care about; the fact that you’re really smart; an argument about a charged political topic; an argument that they should donate their money to something.
This is why I find it exceptionally impressive when an interviewer and interviewee manage to convince me—deeply convince me—that they’re both present. I once thought Ezra Klein the best, or one of the best, of interviewers. I now feel utterly more impressed by what Max Linsky manages to do at his best, on Longform and elsewhere. David Remnick, a famously busy man, once confirmed his presence to Max with two simple words: “I’m here.” But another interview I can’t help returning to—so much so that I listened to it three times in a row one weekend—is his first conversation with George Saunders.
Max and George (I’m going to pretend, for the time being, that I’m on a first-name basis with them) are sitting in the back of a bookstore (what an incredible place to conduct an interview; I can only imagine the scene and it feels breathtaking). George is about to do an event with Ben Stiller. Phones are ringing, apparently because people are trying to get into the event. Max keeps picking up and putting down the phones, and he and George exchange comments about the fact that Max is picking up and putting down the phones. At one point, Max suggests to George that he pick up one of the phones and say, “Hello, this is George Saunders.” George Saunders does not pick up one of the phones and name himself to the caller.
Max and George are physically present—they are literally sitting in a room together. But there’s a much deeper sort of presence that manifests in their interaction. At one point, Max tells George about something a colleague said: “you’d be ashamed to act small around [George].”
“That’s, like, the best compliment I’ve ever heard!” Max exclaims. “How do you take a compliment like that? Do you get better at it? Do you have to get better at it?”
Max really wants to know. You can hear his enthusiasm. He’s spent plenty of time with George, both before and during and (given that George appears on Longform one more time) after the formal constraints of their podcast interview. There’s an energy to their conversation that comes from Max’s admiration for George, and George’s care for his work, and Max being the person he is, and George being the person he is. The same is true of Max’s first conversation with Ariel Levy: she is brilliant and open, and Max cares about how she does her work.
I have been, for many months, trying to figure out how Max does what he does. There is a deep emotive quality to his interviews that makes me feel a human connection through my earphones.1 I can hear the care and joy and pain that the people he interviews go through to do their work, and I can hear his engagement and care. It’s a talent to transmit all of this. I interview people myself sometimes—as a hobby and nowhere near as well as Max does—and I want to understand his sacred art.
I feel that I’ve caught glimpses of what it might be like to interview the way Max does.
When I picked up The Chinese Typewriter, I hadn’t even finished the introduction before I knew I wanted to talk to the book’s author: a historian named Thomas Mullaney. He describes an encounter between Western alphabetic universalisms and Chinese script at the 2008 Beijing Olympic ceremony. I still remember reading his justification that this elaborate project he had embarked on was, in fact, possible:
I do think it is possible to write a history of the Chinese typewriter, and with it the broader history of Chinese technolinguistic modernity, but only to the extent that we abandon any and all fantasies of hearing this machine “on its own terms.” No such aural space exists or has ever existed: no autonomous, unspoilt soundstage waiting to be reconstructed by the historian that, when rediscovered, will redeem the Chinese typewriter by returning it to its rightful place. The aurality of the Chinese typewriter was and has always been a compromised space, at all times somehow related to, completely ensphered within, and yet distinct from the global soundscape of the “real” typewriter of the West.
Who writes like this? I had to know something about the human being who had not only managed to tell a mesmerizing story, but to problematize and justify the place of the historian—of himself—in doing so. I sent off an email, read more, and found myself asking more questions. I discovered Mullaney had been working on this project for over a decade and a half, and, again, I wanted to know why.
I found an answer (or, more accurately, a refusal to answer) in this clip from January 2010, where Mullaney describes the mysterious chemistry between a person and a research project:
got at aspects of this in her wonderful essay on research as leisure. Why do you have the desire to ask and answer the questions you set your mind to? From where do you draw the commitment to evidence, the desire and drive to understand history and produce an artifact?I actually don’t know where the Chinese typewriter project is coming from. I actually don’t know yet why I’m interested in it. That sounds strange, but actually I think that—at least for me—all research originates in some sort of very personal and very self-centered question that one has about the human experience, existing in this universe… I think it comes down to very basic questions that we have about this world and about life. Part of the research project is actually figuring out why this seemingly unrelated thing speaks to you.
I eventually interviewed Mullaney, after listening to every talk and appearance of his I could find, after reading his books twice with pencil in hand, after spilling my guts into another post so I could begin to process what I was reading as I made my way through. There is a possible version of our conversation that spent more time on the content of the books, but instead we spent nearly two hours discussing his understanding of his work, why he was doing it, and how it affected him.
I think the interview worked well mostly because Mullaney was willing to be open—as he had been open in that YouTube interview when he was (in my estimate) around 5 years older than I am now—about the deeply personal nature of his motivations and his connection to his work, and the fact that he hadn’t figured everything out just yet. I asked him what I really wanted to know, and I suspect those questions got at what he really wanted to talk about.
The pull quote I used for that interview really sticks with me:
False universalism freaks me out. It doesn’t freak me out as a first principle because of epistemic violence. It freaks me out because it works.
Mullaney doesn’t just call false universalism bad or account for its risks. He told me that it freaks him out. That’s a very personal thing to say. Again: how do you keep at a project for nearly two decades? And how do you do it with the level of intensity that it seems Mullaney did? There’s something personal here.
Intensity abounds in exhibit 2: in this part of a presentation at the San Francisco Public Library, Mullaney mentions that he went through a number of historical archives—the names of over 10,000 students and their teachers—as if it were a mere matter of course. He didn’t just skim the names; he counted them and found that 60% of these names belonged to women and 40% to men.
I was, and am, fascinated by Mullaney’s dedication to this work and that dedication as an attempt to understand himself. As he explained to me, his obsession with the Chinese typewriter had nothing to do with China, but there was something deeper. This essay from
gives a nice articulation of what could have been Mullaney’s internal monologue:There’s a strange, numinous, beauty to be discovered when you run deeper into the woods instead of backing away at the fringes. If I keep chipping away at this thing, something fundamental will be revealed about me.
For Mullaney, writing The Chinese Typewriter and The Chinese Computer constituted running deeper into the woods. From what he shared, it sounds like what he found pursuing him through the trees was a single, provocative word: disappearance.
I’m still struck by how much Mullaney was willing to share with me. I think a large part of it has to do with the sort of person he is. I hope another part of it was that he could sense I’d given time and care to his work and asked him questions I really, really wanted to know the answers to.
My theory about Max, and a state I think I’ve been able to access myself sometimes, is that his unique ability to listen and engage with his interviewees comes from a place of deep care. Recording an interview with someone that you both know is going to end up on the internet is, when you think about it, really weird. There is almost no way you can interact as you would in normal, everyday conversation; the very nature of what you are doing almost precludes the possibility of really being yourselves.
What Max manages to do is to figure out how to get people to be themselves, and it seems he’s able to do this because he genuinely cares about people. It’s hard to make it in a field like journalism, so it seems plausible that nearly everyone Max interviews is trying to do something they deeply care about. It’s hard not to care about someone like that, and to want them to succeed.2 I think his unique ability to listen comes from this care.3
As it turns out, Max—in traditional Max Linsky style—has sometimes let slip that he, too, is thinking about how to be an interviewer. And, more fortunately, I don’t have to theorize too much about Max, because he’s willing to share details about his craft. He gave a talk, titled “The Art of the (incredibly awkward) Interview.” which is worth watching.
But, even more importantly, he has a clear goal. At the end of one conversation—I no longer remember which—Max tells his interviewee that his goal is to make people think on their feet. He wants his interviewees to join him in a different place from their usual talking points.4 And he’s remarkably effective at this.
Getting someone off their talking points sounds easy: why not just ask a question this person hasn’t heard before? Why not just ask a better question or consider a new connection to something? This turns out to be part of the solution, but not the whole solution. I think I came up with some decent questions for Yoshua Bengio, but certainly didn’t manage to get him off his talking points.
I’ve learned that there are things you can do to counteract this, but you can’t expect them to work all the time. I’ll draw directly from what Max himself suggests: besides “giving a shit” and being well-prepared, ground rules turn out to be a great strategy. But, you have to be very careful with these. Inspired by an early Longform interview with Terry Gross, I tried one of her ground rules for one or two conversations. It went something like: “This is not a live interview. If you say something that you want to correct, or mess up somewhere, feel free to go back to where you started. (and, maybe, start at the beginning of a sentence so I can keep my head when I’m editing this).”
This turned out to work pretty badly for me: one guest who I tried this on, who hadn’t done too many podcasts before, felt more stiff and formal than I’d been hoping for. I think, in her imagination, she had to come across as articulate and poised and very smart. She was all of those things, but I don’t think you need to overdo it in a conversation for people to pick up on this.
All this said, I’ll have to agree with Max that caring and listening (again, assuming you’re well prepared) are at least 80% of it. In some of the best interviews I’ve done, I didn’t have to think about the questions I’d planned to ask or reference notes. Sometimes I did, if I really needed to remember something. But, when things were going really well, the person I was interviewing told me something that made me want to know five other things. All I had to do in the moment was listen to what they were saying, and figure out how to form the incoherent mess of wonder in my head into a question.
But, to re-emphasize it’s much of what you do before the moment that matters: I think you can come to care about someone more by understanding better their commitment to their work; you can listen for the right things in what they’re saying if you’ve engaged with their work. I cared deeply about Tom Mullaney’s project because I could see how deeply he cared. I wouldn’t have been able to ask Tal Linzen about an apparent inconsistency in his work5 if I hadn’t read many of his papers carefully.
This isn’t to say that all interviews will or should turn out deeply personal or reflective. Some interviews are intended to convey information, to give someone an opportunity to speak about their work. I might argue that a talk is a better format for conveying information, but speaking at an audience for even 45 minutes is pretty hard.
That said, whatever the purposes of the weird digital object we call a podcast or a recorded conversation are, people who listen to them can start to crave authenticity. Packaged, easily consumable information is a valuable thing, but I don’t know if that’s the best use of a medium that seems like it can do so much more. Maybe I’m biased. Maybe the fact that I love fiction (or “literature,” whatever we want to call literature) more than anything influences the kinds of interviews I like and what I think I learn from. Max tells interviewers to be willing to sound like an idiot. I think we can go a step further and say we are incoherent and ridiculous, and interviewers who internalize this notion might just convince their interlocutors that there’s something to discover amidst the confusion that is our attempts to make ourselves known and to communicate.
In the same way my months reading Proust taught me more than any number of self-help-y books I spent time with in earlier years, so too do I think conversations that expose confusion and uncertainty and doubt teach us so much more than two-person book talks or paper clubs re-labeled as interviews. I continue to think that the most core truths we can access, the most important things we can know about ourselves and about one another, aren’t expressible as propositions.
This has come up at multiple junctures for me. When I interviewed the neuroscientist Peter Tse, he spoke about how the impossibility of truly understanding certain matters in physics—Everett’s interpretation of the quantum state, that posits many worlds, isn’t falsifiable in the Popperian sense after all—led him to an almost equally interesting field where he could get a grip on something like knowledge.
If there’s any value to the odd media formats in which one person interviews another person, then, I think it’s to leave you—the listener—with more questions than answers. Maybe this is an odd take, given that the whole enterprise is predicated on the act of asking and answering questions, but it feels right to me. They should make you want to know something—hopefully something deeper than the answer to “how much data did this guy use to train the models they’re talking about?”6
I think it’s an interviewer’s job to figure out how to convey just how brilliant and thoughtful the person they’re speaking to is.7 I think this is both hard and effortless. It’s hard for multiple reasons. Part of the goal of preparation is to understand something about the person you’re speaking to. How else can you know what makes them interesting?8 Once you’ve done some work to try to sort that out, there’s a lot you can do: prod, challenge, agree, flounder.
I still remember the first two Longform interviews I listened to: both with Jia Tolentino, whose Trick Mirror had been one of my favorite reads in college. In her second interview, with—you guessed it—Max Linsky, she’s there to promote the book. But the interview hardly feels promotional, except for moments like Jia’s meta-level acknowledgement that she is doing a great job promoting this book of hers.9
About 45 minutes in, Max explains that he can usually distinguish between two drivers when people insert themselves into a piece: they have experiences, and want to use those experiences to explore something; or, an idea comes first, and they can use anecdotes to illustrate the idea. He says that in lots of Jia’s essays, he can’t tell what’s driving.
This leads to one of my favorite parts of their conversation, where Jia says some really interesting things about how she interacts with the world, how she thinks about others’ opinions of her. I listened carefully, because I was on a plane,10 and it was the first time I found myself actively taking notes on a podcast in a long while. My most interesting memory from that interview, perhaps, is Jia’s statement that she never feels she has an original thought—she merely explains what everyone is thinking with extreme clarity.11 That’s not the kind of admission I hear people make very often. It only gets more complicated and reflective from there, and it’s beautiful to hear.
I sometimes have mixed feelings about the podcast medium, because I think its accessibility to producers can demean its stature: you don’t have to work terribly hard to put something out there, and, sometimes to your chagrin, you can just say stuff. But the form has its virtues. I think interviewers like Max illustrate what it can be at its best—and I want what that gives us. It, like any other medium, is capable of giving us something we should cherish deeply: something that does not demand our attention, but is worthy of it.
In keeping with the theme of discovering more questions than answers, it feels fitting to end this essay with a few of the questions Jia asks herself, Max, and us. The best words she’s able to find for Trick Mirror’s subject are “the impossibility of knowledge.” Her questions express deep complications not just in the pursuit of knowledge, but in doing almost anything at all.
Can you be sharp and uncertain at the same time?
Can you generate something worthwhile in a space of deep uncertainty?
Can you generate hope against a background of deep futility?
Can you access morality in an atmosphere of immorality?
What does any of this look like?
Jia posits that you can find clarity through uncertainty. Another theory is that they’re the same thing—confused clarity might be all there is. She and Max seemed to find that confused clarity. Maybe I can stammer, erase, and fumble my way there too.
Listen to this interview Max did with Sarah Schweitzer. It’s wonderful. The dialogue is wonderful. Max is so charming. Max is so charming! How do you become that charming? I wish I could be that charming.
Everything I’m saying assumes that an interviewer is doing their level best to be extremely prepared for an interview, which hopefully doesn’t need to be said. I think you need to do this to allow someone to open up, because it allows you to understand something about them and what they care about. While it wasn’t a particularly personal conversation, I spent many hours reading about variational (quantum) Monte Carlo methods and wave function ansatz so I could get a grip on David Pfau’s papers and not sound like an idiot when I eventually spoke with him. I think this sort of preparation also allowed my conversation with Ken Liu, one of my favorite science fiction authors, to be incredibly generative.
When I say Max is a good listener, I mean this emphatically. There are good listeners, and there are good listeners. It’s an incredible joy to come across one of the latter—but you have to listen for them.
Of course, when you’re talking to someone about their work, you can’t do this at every moment! You do have to establish some context about what it is you’re discussing, after all. But something you can do is spend less time on the details—maybe don’t have someone recount the entire details of a story to you. You can tell your audience, “You should go and read the piece. This will be a lot more enjoyable if you do.” And the Longform hosts have often done this. When I spoke with Joss Fong, she convinced me that something like this is actually a great way to deliver something useful to people: if you don’t demand anything from your audience, there is only so much you can deliver. If you are willing to expect that your audience will look things up, read, and do a little bit of work to engage with what it is you’re trying to offer, you can give them so much more even if the size if your audience is smaller.
In “How to Plant Trees in LMs: Data and Architectural Effects on the Emergence of Syntactic Inductive Biases” he asks the (loaded) question: why does simpler language teach syntax more effectively? In an earlier paper, “Assessing the Ability of LSTMs to Learn Syntax-Sensitive Dependencies,” he claims neural networks can be encouraged to develop more sophisticated generalizations by oversampling grammatically challenging training sentences. This isn’t a contradiction, but there is a tension.
Recently, I had about an hour with Peter Lee. I used 20 of those minutes to ask him about compiler generation and type theory. When he told me that, in his work, beauty and utility didn’t come apart (the very problems he found beautiful turned out to be practically useful—how lucky!), I wanted to know what beauty looked like for him so that this was true. So I asked him. He told me it almost felt like we were discussing how many angels could dance on the head of a pin. This exchange felt a little awkward, and I think that was good. Awkwardness can be good.
Note that this also assumes you’re interviewing a person because you actually want to talk to them and not because you have to release an episode this week and this is what you could figure out. That feels like a waste of everyone’s time. You might sometimes have to do this if interviewing is your livelihood, but that’s not my concern.
To re-emphasize, this is different from knowing that someone is interesting. I think the “why?” question is also a lot harder. If I had to put this into Mullaney’s terms: studying someone’s work and trying to understand them better is also an interrogation of why you, personally, find this person interesting.
At another point in the interview, Jia asks “Have you ever been on Splash Mountain?” to make an analogy to writing. Max answers, “Sure,” in that noncommittal “Maybe I’ve been on splash mountain, and maybe not, but I’ve experienced something similar enough to what you’re asking about that I’ll understand the analogy” kind of way. So, I don’t know if Max has actually been on Splash Mountain. It would be neat to have a chance to ask him one day.
I wasn’t hardcore enough to rawdog that particular flight.
Apparently, The Paris Review will call your essays “original” even if you think the ideas in them are not. This interview and her interview with Max Linsky came out the same day, and she says a few of the same things: that she knew the question she wanted to ask in each essay, for instance. But she doesn’t seem to divulge as much to Brian Ransom as she does to Max. The questioning doesn’t quite take the conversation to the same place—I’d guess it’s somewhat influenced by the interview’s fate to appear as a print piece.
1. On dedication to work as an attempt to understand oneself: This is a theme I think tackled well in the film 'Adaptation' (Charlie Kaufman). What is elaborated in that film is that obsession has this recursive, gravitational pull: a journalist becomes obsessed with deconstructing the root of obsession of an orchid thief; she craves deeply to know and understand what it feels like to be obsessed with something. A film maker becomes obsessed with his portrayal of the journalist and the orchid thief. We cannot see the end to the recursion which is what it makes its gravity so strong. We are all achingly looking for something to get lost in (to disappear, as you say), and if we are lucky, find the looking is what gets us lost.
2. Re this: "Max tells his interviewee that his goal is to make people think on their feet. He wants hits interviewees to join him in a different place from their usual talking points.4 And he’s remarkably effective at this." - Henrik's essay on Werner Herzog's interview style elaborates on this point well:
///
The interview starts with the Reverend talking about how the Lord works in mysterious ways, and so on—polished clichés. You notice that Herzog is trying different prompts to get him to open up. Nothing works. The Reveren keeps repeating things he’s said before—words that protect him from his thoughts, that allow him to perform the role of Reverend, instead of being what he is: a man named Lopez, who will soon have to feel a man go stiff under his hand.
At one point, the Reverend Lopez, as a part of a monologue about the beauty of God’s creation, mentions that he sometimes meditates on the beauty of the squirrels he sees on the golf course. Herzog says: “Please describe an encounter with a squirrel.”
Lopez is a bit surprised by the question, but he takes it in a playful spirit—his voice lifts, joyously. He starts to talk faster. He is no longer saying versions of things he has said before, he’s not a facade, he’s just there. The entire interview has changed into an open human drama.
///
Great read! Doing 70+ interviews will certainly result in having some insight into this stuff.
I think a lot of this can be boiled down to saying an interview should feel like a real conversation that is interesting to listen in on even while you cannot participate in it. If the interviewer is mostly absent, it's no longer really an interview, is it? But there are nuances here - who are your hypothetical listeners? What will they find interesting?
That's why I think the idea that "Everything I’m saying assumes that an interviewer is doing their level best to be extremely prepared for an interview, which hopefully doesn’t need to be said" should not always be true. If you really want to do a super deep dive into the details of someone's work, yes, this make sense. But if you want to provide an overview, going in with surface level knowledge and just a rough plan of attack can be good, in the sense that you are put in the position of the listener, and must actually think on your feet instead of trying to queue up the next question you've thought up before.
At the end, it's more art than science, even if there are rules of thumb and some wisdom that can be garnered by listening to the greats.