when two people talk, a third thing emerges

observations on generating great conversations

Did someone forward this to you? Subscribe to get future essays directly in your inbox.

Claude Monet, The Beach at Trouville

I used to be terrible at conversations. God-awful. The kind of person who, if asked “how are you?”, would either clam up or launch into a detailed explanation of my emotional landscape, complete with footnotes and asides. It’s odd because, as a kid, I could entice even walls into full-fledged conversations. It was as if, somewhere between getting braces and nearly failing a maths exam, someone entered my brain and pulled down a Big Red Lever.

I’m happy to report that I am not that person anymore—I have managed to pull the Big Red Lever back to a somewhat midway point, if not entirely up. But it took a lot of work. It took a lot of asking great conversationalists what their secret was, and getting frustrated when they replied, “just talk to people, Sindhu”. A lot of tiny experiments in anxiety-inducing social settings, at work, on sales calls. A lot of realisations that, contrary to popular belief, I actually enjoyed human connection, but had to figure out the mechanics of it because it didn’t come naturally to me.

I can feel a longer essay bubbling up inside of me—mostly because I can’t resist the pull of more research! More curiosity!—but until I put that together, here are some of my observations on the art and science of making conversation.

While making conversation, we tend to default to what I call the ‘lowest common denominator’ rule: trying to offend the most offendable person. Often, that means playing it safe near the median line, not being weird, and being over-cautious of crossing boundaries (even though we don’t really know where they end). It’s an un-negotiable, one-size-fits-all approach. This can work if, for example, you’re passing time or not too interested in establishing a deeper conversation. On the other hand, this is a death trap if you’re actively trying to build a rapport with someone. It congeals into an act of circling each other, waiting for the other person to make the first move. 

Instead, it’s a lot more fruitful if you aim to elicit more interesting answers to questions you ask. Gently, respectfully, probe the edges of comfort. Make a beeline for the edge of what this person has ever been asked before, and you might find that people have the capacity to be a lot more interesting than when they’re drawing on a bank of templatised answers. 

In her 2016 book on reclaiming conversation, sociologist Sherry Turkle claimed that it takes seven minutes for a conversation to get really good. I’d wager that you could reach this point with 2-3 really good questions (I’ve written more about what makes a good question here). 

Take the classic “Where are you from?” exchange. It’s safe, it’s polite, but it has potential that we might end up not tapping into. You could accept it when the other person says “Bangalore” and move on. Or, you could aim for the edge: “What’s something about Bangalore that would surprise most people?” That usually gets you past the standard traffic jokes and straight into more interesting territory. You could follow that up with, “If you could take one thing from Bangalore and make it mandatory everywhere else, what would it be?” Now we’re talking.

A mistake I’ve seen—and made all too often—is sharing too much personal information too quickly in the name of being truthful and open. Emotional honesty is good, but bringing up more personal or private details too early on in the conversation can create a sense of obligation in the other person, that they also have to share something equally vulnerable even if they’re not comfortable with it. It can also catch them off guard: they might not know how to reply or, worse, might say something that instantly sours the conversation. Meaningful conversation needs a build-up; it’s less about immediate depth of disclosure and more about a graduate pooling of trust. Ironically, sharing everything at once just short-circuits the possibility of a deeper connection. 

You can only ask so many questions in a row before you start interviewing the other person. The canned advice for making better conversation is “be a good listener”, but often that’s just translated into ‘ask more questions and let the person respond’. It may feel like when you’re asking a lot of questions, you’re displaying curiosity about the other person and generously giving them the limelight. But this approach can be unkind: it offloads all the effort of coming up conversation points and directions to the other person. It can be exhausting for them to keep giving without receiving anything in return (unless they love the sound of their own voice, in which case, carry on). Of course, you shouldn’t shoe-horn your way into the conversation, either, or be too eager for your turn to talk. 

The key is finding the sweet spot between interrogation and conversational hijacking. This one took a bit of trial and error, but a 2-1-1 approach, in my experience, is a safe bet to bring up a related experience of yours at the right time. I mean: ask a couple of questions, share one relevant experience of your own, end with another question, and see how your partner responds.

Say you ask someone what their startup does, and they reply with: “We're building an AI-powered tool that helps small restaurants optimise their food ordering to reduce waste.” You might ask an engaged follow-up: “That's fascinating! What made you focus on small restaurants specifically?” to which they might reply “Well, the big chains already have sophisticated systems, but smaller places often can't afford them. Plus, I love the personal connection you get with local restaurant owners.”

By now you’ve gone back and forth twice, so you can insert a bit of your lore: “I really connect with that mission. I did some consulting work with a family-owned Vietnamese restaurant, and it was eye-opening to see how thin their margins were and how much every bit of waste affected them. Have you started piloting with any restaurants yet?” From here, it might go two ways. The other person might respond to your question. But if they’re also a good conversationalist, they’ll probably respond to your question and then come back to your experience: “Have you kept in touch with the restaurant owners you consulted for? I'd love to hear what other challenges you noticed in their operations.” Either way, you’ve balanced out the conversation. 

I fully realise this is only the tip of the iceberg that is the art of making conversation. So much of it can’t be abstracted or templatised, for good reason, I think. But I think the bottom line is that to make good conversation, you have to make it—you can’t magic it out of thin air. You have to lead with curiosity and intention. You have to want that conversation, and that means being an equal and willing participant. 

I read somewhere that, in a conversation, you aren’t only exchanging, you’re creating a third thing. I really like this, because it frames conversations as a shared act of curiosity. The topic becomes almost secondary to this process of mutual discovery. I also think this is why truly engaging conversations can’t entirely be scripted or reverse-engineered once they take on a rhythm of their own. This is a rare Pokémon, though, which would explain why when asked, we can only think of a handful of conversations we’ve had that we might describe as illuminating or perspective-altering.

A quick note: I recently launched an independent digital publication called Patina! It will document what I call stories of living culture: craft, culture and community as they persist and involve in food, art, ritual, objects and heritage. Submissions open in April 2025 and the inaugural volume will be out in June 💌

Read more about Patina on its website, readpatina.com. Writers, photographers, documentarians, readers, please sign up for pitch calls and theme updates!

A big thank you to all subscribers supporting my work. If you’d like to support me financially, please consider signing up for a paid subscription at just INR 149 (~$1.75) a month or INR 1499 (~$18) a year. Other ways to help out is to like my essays on Substack, subscribe for future essays (if you’ve come from a shared link), and share my essays with anyone you think will find them valuable.