Conversation, Interintellect, and Arcadia (with Anna Gat)
What this covers
Anna Gát, poet, screenwriter, and founder of Interintellect, joins Russ Roberts to argue that structured conversation—far from being killed by technology—can be revived through deliberate craft. The conversation rests on a distinction between two kinds of knowledge: substantive works like poetry and music, which are fixed entities whose loss resembles losing a person, and instrumental knowledge like science, which progresses, can be debunked, and improves collaboratively toward external truth. Gát contends that conversations need hosts—figures of temporary authority who listen, reflect, and synthesize—not to constrain participants but to free them. She argues that Stoppard's Arcadia dramatizes this distinction while exploring how scientific progress and human attraction complicate each other, and she connects the play's themes directly to her mission of preserving the practice of thinking together across generations.
The episode ranges across how conversations actually work: why a third thing)—a book, topic, or idea—and three or more people enable breakthroughs that a two-person dialogue cannot; how physical positioning affects difficult exchanges; why explicit framing of conversational roles ("do you want advice or listening?") undermines authentic interaction; and how hosting maps onto a hierarchy of facilitation, moderation, and mediation. Gát and Roberts examine why James Gleick's Chaos directly inspired both Arcadia and a feature film, making it perhaps the most influential science book for drama since Darwin. The conversation also contests whether artificial intelligence can replace human coordination and whether academics across disciplines face real incentive structures that prevent synthesis. Throughout, the pair treat Arcadia's plot and characters not as the author's pronouncements but as dramatic expressions, following Mamet's principle that every line must answer who wants what from whom.
Gat argues that conversation is a craft that can be deliberately structured—through hosts, authority, formats, and a 'third thing'—so that technology, far from killing dialogue, can revive the salon; and that Stoppard's Arcadia dramatizes the deeper distinction between instrumental knowledge (which progresses and can be debunked) and substantive human creations (which, like lost people, are timeless and irreplaceable).
- Conversations are not democratic; a host who absorbs the listening and exercises temporary authority makes participants freer, not more constrained.
- A 'third thing'—a book, idea, or topic—and group size larger than two are what allow real breakthroughs in conversation.
- Arcadia distinguishes instrumental knowledge (science, debunkable, progressive) from substantive works (Bach, Byron) that are fixed entities whose loss is like the loss of a person.
Hosting and conversational structure enable genuine human connection and breakthrough thinking by managing power dynamics and creating safe space.
- People feel freer to discuss the great questions of life—spirituality, morality—when they approach them through a 'third thing' such as a great book, film, or scientific discovery; the seemingly external vessel relaxes participants into the deeper conversation.
“You will feel more liberated to talk about the great questions of life, spirituality, morality. It's about Dostoevsky and not about spirituality and morality.”
- Someone attempting to dominate a conversation is often trying to take over the hosting role without being the host, a reaction to a power vacuum created when the host is not present enough or has failed to shift from facilitation to moderation.
“I often find that somebody trying to dominate is a way to take over the hosting role without being the host. And, it's often a reaction to a kind of power vacuum.”
- Hosting maps onto a priority hierarchy analogous to pilots' 'aviate, navigate, communicate': facilitate (stay light, ensure everyone gets a turn), moderate (intervene when real disagreement or a domineering person appears), and mediate (resolve conflicts after they erupt)—and a well-run space rarely reaches the mediate stage.
“You have Facilitate, Moderate, Mediate. Facilitate--in the word, it has the word easy, right? It's chill.”
- Explicitly naming a conversational role—as when Americans ask 'do you want me to listen or give advice?'—is undesirable because it removes the veil of human behavior; one chooses a format only to enter the real situation, after which the seams should disappear and it becomes real.
“it kind of removes the veil of Maya, of human behavior. You don't want to feel that you're acting... You just choose a format at the beginning as a way to enter the real human situation, but you're not going to hold it throughout”
- Conversations are not democratic; one of the best ways to defuse opposition is to have a host—any accepted figure of temporary authority (a mother, the person who booked the restaurant)—who does the listening that participants cannot, reflects back what each person said, and synthesizes opposing views into a dialectical third place that moves the discussion forward.
“Having people be able to trust somebody and outsource the listening to this person as a host, my greatest task on Interintellect... is I'm listening because people don't want to or cannot.”
- A two-person dialogue is inherently tense and tends to loop back on itself because there is no third element to break the impasse, whereas adding a third—reflected in the word 'tribe' deriving from 'three'—lets a breakthrough or new synthesis emerge.
“It's very difficult for two people to do... When you are the two of you, you get back to the same things because there's no third to take you out.”
- Because humans are animals, a head-on, face-to-face physical setup is read by the 'monkey brain' as a fighting position, so difficult exchanges (like a doctor delivering a hard diagnosis) go better when people are positioned diagonally rather than across a desk, which is why friends naturally arrange themselves at angles.
“it's actually really bad to have this facing each other across the desk set up because your monkey brain feels that it's some kind of opposition”
Substantive works embody irreplaceable knowledge whose value lies within themselves, distinct from instrumental knowledge that serves external truth.
- Human knowledge divides into substantive works (a Byron poem, a Bach piece, Hamlet) that are fixed entities whose telos lies within themselves and whose loss is as tragic as losing a person, versus instrumental knowledge (scientific laws, historical biography) whose telos is an external truth, which can be collaboratively improved, debunked, and replaced as merely getting closer to the goal—so the real distinction in Arcadia is not science versus poetry but substantive versus instrumental knowledge.
“there really is nothing to improve on a Byron poem. You listen to Bach and you don't think that somebody could come and debunk Bach.”
- Plays cannot be spoiled the way crime TV series can, because one does not read or watch a play like Arcadia or Romeo and Juliet for its plot—the outcome is often announced in advance—but for its language, ideas, and execution.
“you can't spoil plays. Plays are not like crime TV series. You know how Romeo and Juliet stars and they tell you they will die. We can totally spoil Arcadia. You don't read Arcadia for the plot.”
- In dramatic writing nothing a character says is the author speaking directly; following Mamet, one must always ask of any line 'who wants what from whom, what happens if they don't get it, and why now'—so Septimus's speech on lost knowledge is a character coping emotionally with loss, descriptive rather than a normative truth Stoppard pronounces.
“it's a play. It's never Stoppard talking to you. If he wants to talk to you, he will write an essay”
- Because we have no idea how poets actually worked and letters get lost or burned, any account of a work's genesis can be overturned overnight by a new discovery—taking careers and tenures down with it—mirroring how a scientific finding is only valid until someone proves it wrong, so a scientist's goal is to grow old and win awards before that happens.
“we have no idea how poets worked, letters get lost, like, things burn, and whatever you're reading in literary history, somebody can always come by tomorrow and find something else”
Academic and professional institutions create perverse incentives that penalize interdisciplinary work and reward competitive rather than collaborative truth-seeking.
- In academia the humanities and sciences function like rival football teams with strong incentives never to settle, so people who collaborate across disciplines and create interdisciplinary dialogue get ostracized by both communities and face little career upside for a more synthetic approach.
“People who actually collaborate and create interdisciplinary dialogue and research get a little bit ostracized by both communities. So there is little kind of career upside also to try to go at it in a more synthetic way.”
- Arcadia dramatizes two kinds of academic: Hannah, who values the relentless pursuit of knowledge for its own sake and almost cannot care where it leads, and Bernard, who treats research as a sport with winners and losers and wants the truth chiefly so as not to be debunked faster—even squeezing inconvenient facts into his theory to tell a compelling story and win publicity.
“you have Hannah who... says that it's about us trying. It's about the kind of relentless pursuit of knowledge. You almost can't care about where it's going.”
- Playwrights and screenwriters since Aristotle have understood how to write dialogue almost as a science, working in formulaic ways that enable creativity—yet we have never systematically tried to help real human beings have conversations with each other the way we help fictional characters.
“it never made sense to me why, other than maybe Gottman's science of couples therapy and other fringe experiments, we never really tried to help real human beings have conversations with each other the way we are so adamant of helping completely fictional people to have great conversations with each other”
Mastery and talent are universal, yet isolation and systemic blindness prevent recognition of genius in those nearest to us.
- Every performer—teacher, actor, presenter, dancer—follows an arc in which 'with great routine comes great routine': mastery makes it hard to be surprised and equally hard to be genuinely impassioned, even as one's craft improves.
“I think every performer, whether that's a teacher or an actor, a presenter, a dancer, will have this strange arc where with great routine comes great routine.”
- Talent is ubiquitous, and the bigger tragedy is not that talent is destroyed or misunderstood but that it is missed entirely by people standing right next to it—so one should look around one's own dinner table rather than hunt for talent elsewhere, just as Lady Croom fails to recognize her own daughter's genius.
“it's much less about going out and hunting for talent and more around looking around your own dinner table... You have no idea what this person is working on in private”
Human coordination problems and emotional bonds remain constant despite technological change and geopolitical upheaval.
- Despite enormous cultural and geopolitical upheaval over the past eleven years—Brexit, Trump, a pandemic, wars—the basic human need for one another never changes; AI will be an enormous break but cannot raise the barn, raise your children, or have the kind of conversation in which two people reach a new level of self-understanding, because so much of life is a coordination problem.
“humans really do need each other. And, they need friends of a different generation; men need women, women need men, scientists need poets, poets need scientists. We can't really get much done without each other.”
- The one thing science has no resistance to is attraction between people, which messes up all the models and systems and complicates everything—dramatized in Arcadia by the wife who would upturn the Newtonian system.
“one thing science has no resistance to is attraction between people, which just messes up all the models and all the systems and complicates everything”
- Every generation since biblical times has believed the world was falling apart and that it is the last generation, because the idea of collective continuation after one's death is somehow less comforting than the thought that everyone else dies too.
“Every generation since biblical times has thought that the world was falling apart and that this is the last generation... because somehow that's more comforting to us than the idea of our own death.”
Conceptual frameworks must exist before discoveries can occur, explaining why innovations emerge when they do and not before.
- Discoveries cannot precede the conceptual structures that make them possible—as Valentine argues, Thomasina could not have innovated mathematically the way 1990s researchers imagine, because 'you cannot open a door till there is a house'—which explains why inventions arrive when they do and not earlier.
“she can't have known: You cannot open a door till there is a house.”
Interintellect's pricing and intentional design model produces focused participants and eliminates toxic conversational dynamics at scale.
- Toxic and domineering behavior occurs surprisingly rarely at Interintellect because nothing in the format incentivizes it: people must pay and intentionally allocate time to attend, producing well-focused, well-behaved participants, and tens of thousands of events over six and a half years have run without toxic incidents.
“It happens surprisingly rarely because it is set up in a way that nothing incentivizes that kind of behavior. So, first of all, people pay to come to Interintellect events”
- Anna Gát started Interintellect in 2014, eleven years before the recording, while studying theater for her third master's degree in London.
“I started Interintellect altogether in 2014, so 11 years ago. I was studying theater for my third master's degree--because I make weird life choices--in London.”
Chaos theory inspired major literary works, demonstrating science's capacity to shape dramatic imagination and artistic innovation.
- James Gleick's 1988 book Chaos, which explained chaos theory to a general audience, directly inspired two extraordinary works of literature—Stoppard's Arcadia (his first play in eight years) and Mike Leigh's film Naked, in which a character reads the book on screen—making Chaos arguably the most influential science book for dramatic writing since Darwin.
“he writes this book and it leads to two extraordinary works of literature, basically. One is Arcadia... And, Mike Leigh also reads it and it leads to the movie Naked”