What science still can’t explain about love

Tennesha Wood is a professional matchmaker but admits she can’t perfectly predict when a couple will hit it off. “People give me really specific descriptions,” Wood says. “They might say, ‘I want a 6-foot-tall Black man with nice teeth and a bald head.’”

Wood is the owner of The Broomlist, a matchmaker agency for Black professionals, and her clients will often tell her exactly whom they’re looking for. She’ll set up a date with someone who fits the description. “And they’re like, ‘Yeah, I just didn’t like that 6-foot-tall Black man,’” she says, laughing.

Matchmaking involves a lot of skills you might find in a therapist — sharp questions, active listening, steady coaching — but it is not a perfect science. “I guarantee I’ll introduce them to somebody that does fit the things that they’re looking for,” Wood says. “One thing I cannot predict and make happen is that initial chemistry.”

This is one of the great mysteries of life, and there’s a scientific sub-discipline devoted to it: relationship psychology. I started reporting this week’s episode of Unexplainable, Vox’s science podcast on unanswered questions, by simply asking relationship researchers: What don’t you understand about love?

It turns out they grapple with the same question as matchmakers, romance authors, poets, and many others. “The big mystery is — do you really know who you want?” says Dan Conroy-Beam, a University of California Santa Barbara psychologist who studies relationship formation.

The question seems simple, but it’s not trivial. A lot of time, energy, and heartache goes into finding solid relationships. “In a lot of senses, who you choose as a partner is the most important decision you’ll ever make,” Conroy-Beam says. “That’s going to affect your happiness, your health, and your overall well-being.”

Scientists don’t have all the answers, and they often disagree on which answers are even possible. But I found that their hypotheses — along with some advice from matchmakers and relationship coaches — can help us think through how love starts and how to maintain it once it’s found.

In the 20th century, Conroy-Beam explains, scientists in his field kind of just trusted people’s stated preferences — for example, how tall, funny, or attractive they hoped their partner would be. The assumption was based on thin evidence, though. A questionnaire can ask participants to rate hypothetical dates, but that doesn’t tell you how the date is going to go.

In the 2000s, “scientists in this field were really starting to wonder, like, how useful this data was in the real world,” Conroy-Beam says. “So researchers turned to something that was pretty popular at the time, which was speed dating.”

Invented by a Los Angeles rabbi in the late ’90s, speed dating was kind of like Tinder in physical spaces — a way for singles to meet a lot of other singles and make snap judgments about them. Psychologists took notice of the trend and swiped right, so to speak. Speed dating is “a really nice invention for psychologists,” Conroy-Beam says, because it’s essentially an experiment that generates lots of data.

Psychologists started to publish studies. Before the dates, researchers could ask participants what they wanted in a partner; after, they could compare those notes with who they actually chose to go on more dates with. And then they could see: Was there any pattern?

“The answer turned out to mostly be no,” Conroy-Beam says, “much to a lot of people’s surprise.”

People who go on dates tend to make guesses based on what they like, says Paul Eastwick, a psychologist at the University of California Davis who also studies relationships. We might think, “I click really well with people who are interested in anime or people who are really interested in vegetarian cooking,” Eastwick explains. “The issue is that we really can’t find any evidence that any of those kinds of factors matter in terms of matching people.”

These speed dating studies have limitations. Even though they were set in the “real world,” outside of the lab, speed dating is just one of many ways that people meet. These studies, and much of the research discussed in this article, are also built on Western and American assumptions about dating. To grossly oversimplify, dating in the US tends to be individualistic (compared to cultures where families might play broader roles in matchmaking, for example).

“The research is also mega-dominated by cisgender, heterosexual people,” Conroy-Beam adds. “And there’s very little research on LGBTQ people or relationships, and that’s also something that really needs to change.”

Those reservations aside, the results of these studies did cause researchers to reevaluate their assumptions about how relationships form, and it split researchers into roughly two camps.

Broadly speaking, one group argues that compatibility is predictable or follows some patterns. The implication is that love is something we can find. The other group argues that love develops unpredictably, even chaotically — which also suggests that it can be built between people who don’t expect to be compatible.

“I’ve heard people say, ‘Attraction is like an earthquake. You just can’t predict when it’s going to happen in advance. It’s inherently unpredictable,” Conroy-Beam explains. “Then there are some people — and this is a camp that I tend to fall into — that think people probably do roughly know what they want.”

Conroy-Beam still believes that people’s stated preferences can help predict whom they’d hit it off with. Our preferences might not match up perfectly with whom we decide to date in practice, but he thinks those preferences are still part of the mental software that guides us to a match. He suspects that as scientists look more closely, they will discover those preferences at work.

When we’re making dating decisions, Conroy-Beam says, most people aren’t filling out a mental scorecard. It’s not like we’re thinking to ourselves, well, this person scores six on attractiveness and a seven on humor, which adds up to a passing score and a second date. He thinks it’s more complicated and involves trade-offs. For instance, you might really want to be with someone who’s really smart, but you wouldn’t want to be with them if they were also really arrogant.

Wood sees this dynamic a lot. Her clients might say, “I want somebody who is educated, athletic, attractive, has really good family values,” Wood explains. “Let’s say it’s the case of a woman. I find her a man with all these qualities. I put her in front of the man. Let’s say the man is 5-foot-8 — and immediately, all of those other qualities she wanted seem to be less important because of this.”

When you make trade-offs in practice, you end up with someone who doesn’t look like your dream date, but your preferences got you there nonetheless. The system grows more complicated when you consider that everyone else is also processing their list of trade-offs. Conroy-Beam uses the term “mating market,” as if to suggest that we’re all buyers and sellers and each date is a negotiation. You can’t just choose a partner; they have to choose you back.

This hypothesis is hard to test in the real world, so for now, Conroy-Beam is trying to model it in his computer. His work is almost like a science fiction thought experiment come to life: Let’s say you take real-life happy couples, wipe their memories of ever meeting one another, and put them back into the world. If they meet again, do they hit it off? Is the love discovered again? If preferences matter and guide our decision-making, then there’s a good chance that the amnesiac love birds will find each other again. (If this reminds you of a Black Mirror episode, you’re not wrong.)

Conroy-Beam can’t wipe the minds of his study participants, but he can create mind-wiped versions of them in a computer. He first asks real-world couples lots of questions, individually — what they want in a partner, what their actual partner is like. “Once we have that information quantified, we can create a little simulated representation of you inside of our computers — avatars — that want all the same things as you have and also have all the same characteristics as you.”

He then puts these avatars in a computer program with other couples who have had their memories wiped. And then gets them flirting. “We can see what kinds of decisions actually do a good job of putting people back with their real-world partners,” he says. The idea is that, if he can craft a model that recreates something that exists in the real world, it will probably be onto something important.

In his best effort, his models put around 45 percent of the couples back together, and he says the couples that are put back together in the computer tend to be the happier ones in the real world. That gives him some hope that his models can lead to better predictions of who will hit it off with whom.

However, it’s one thing to recreate a couple that already exists in a computer simulation. It’s much harder to predict couples that don’t exist yet. That’s the next step. “We’re hoping to use our algorithm to take single people, run them through our simulations, and make recommendations,” he says. “We’ll see how well that does.”

Eastwick, the UC Davis psychologist, has a very different take. He doesn’t think it will ever be possible to accurately predict couples before they form. “It is very, very hard to study relationships before two people will officially call themselves a couple,” he says. It’s just too chaotic of a system.

He suspects that a lot of the course of an early relationship is the product of chance. In a chaotic system, small changes in starting conditions can lead to widely divergent paths later on.

When you’re looking at a happy couple, he says, it’s like looking at a chessboard in a game that’s 16 moves in. “Maybe a master could have predicted [the position of the pieces] from the first move, but most people can’t,” he says. There are often many paths the game can take to get to the same position. “It’s worth having some humility about the role of luck and chance in getting this couple to this point,” he says.

Starting a relationship is a process of saying yes to a series of choices: Do you want to go out on a date? Do you want to do this again? Bowling or movies? If your first date is ruined because a bad night of sleep leaves you grumpy and unreceptive to your date’s jokes, maybe you won’t bother going on a second.

“We’re pretty bad at studying unfolding choices over time, setting people on a path to something good or a path to something bad,” Eastwick says. “And it’s largely because we don’t have the data. We weren’t there in the beginning when you decided to start dating each other after barely knowing each other. ... These tools, we don’t exactly have.”

In Eastwick’s mind, the answer to Conroy-Beam’s thought experiment — would couples with their minds wiped find each other again? — is no.

“Take a happy couple and you wipe their minds, and there is a very good chance that you would get a very different outcome,” he says. “There is nothing about the truth of those two people, separately from each other, that does a very good job at predicting where they’re going to end up. It was about choices that they made along the way and the other chance circumstances.”

For Eastwick, the more interesting research question is less what gets a couple together, and more what keeps them together. “Compatibility comes from sort of a series of stacked-up choices that can’t be easily unwound,” he says.

Eastwick believes that love isn’t discovered between two people but grown. He suspects it has to do with setting up a “groove,” or patterns of behavior that reinforce the relationship. A groove can be support for one another’s career ambitions or nightly dinners together to reconnect after work, or something else — and what works in one relationship might not work for another.

He doesn’t pretend to know the secrets of compatibility. “We really have almost no ability to explain any of it,” Eastwick says. “It’s like the dark matter equivalent in relationship research. ... Where does compatibility come from? If it’s not about you and them, it has to be coming from something that is created along the way.”

Psychologists are hardly the only experts on love, so I was curious to ask matchmakers and relationship coaches this same question: Do people really know what they want? Are relationships discovered, or are they built?

They see a bit of both. Yes, people have a vague sense of what they want. But they owe it to themselves to investigate those desires further. “I think a lot of times people do not have clarity on what they want,” relationship coach Damona Hoffman says. But they’ll have the start of it. They’ll have a preference that needs to be unpacked further to reveal a core value they want to share in a partner.

Hoffman’s clients often say something like, “I met this person. How do I know if it’s going to work out?” In response, she says, “I can’t predict whether it’s going to work out, but I can tell you how to get clarity in your goals.”

For example, Hoffman says a client will say, “I have to be with someone Jewish.” Why? If the client isn’t particularly religious, maybe they should consider whether it’s more about living up to a family’s expectations. “Then we go to the root of that,” she says. “Why do your parents feel that that is important, and is that a value that you still hold?”

If it’s about wanting to pass on a tradition to eventual children, that’s a helpful core value to identify, too — and it might lead to a different preference than “I want someone who is Jewish.”

The relationship experts I spoke to agreed that initial chemistry is hard to predict. But they told me that, ultimately, it might not matter as much as these shared values.

“It has a huge element of luck and chance,” Hannah Orenstein, a former matchmaker and current dating journalist, says of forming relationships. “I think if you observe two people on a first date, I don’t think that you can predict long-term success at all, because you’re just scratching the surface of these people. You can’t get an entire person’s life story and values and what they’re looking for in an hour or two hours. But I think if you track that couple for six months, I think you would have a pretty good understanding.”

If there’s anything that everyone agrees on, it’s that a good relationship takes time. “There’s no way I can guarantee feeling butterflies, but I always encourage people to go on a second date,” Wood says. “If you feel like this person shares your values and you have the things in common that are important to you, give it a second date.”

Meradith Hoddinott contributed reporting. Listen to the full Unexplainable episode on Spotify, Apple Podcasts, or wherever you find podcasts.