Love languages: where the idea came from and what research found

A pastor’s counseling notes became a book

Chapman has spent most of his working life in a church. He joined the staff of Calvary Baptist Church in Winston-Salem, North Carolina, in 1971 and stayed about fifty years. His doctorate is in adult education, and he’s a pastor who counsels couples, not a psychologist.

In his office he kept hearing one story in different forms. A husband cooks, cleans and pays the bills, and is sure he’s doing his share; his wife feels distant because the two of them never find time to talk. Chapman says he sat down and read through twelve years of his session notes to look for patterns, then tried what he found on individual couples and later in small groups, where the results looked similar. At some point he decided to write it up for general readers, without psychological jargon.

Another couple, married thirty years, came in with a variation: the husband did a great deal around the house, and the wife still didn’t feel loved. The chores weren’t answering anything she needed. As Chapman put it, what makes one person feel loved doesn’t do it for another.

The idea started at home as well. By Chapman’s telling, his wife once asked why he didn’t help around the house if he loved her, and he began doing chores. His own language is words of affirmation and hers is acts of service, so he has vacuumed for years, though he doesn’t enjoy it.

The Five Love Languages: How to Express Heartfelt Commitment to Your Mate came out in 1992 from Northfield Publishing, an imprint of the Moody publishing house. Moody had turned down an earlier manuscript of his in the 1970s. Sales began modestly: about 8,500 copies in the first year, roughly four times what the publisher had expected, and 137,000 after two years. The book first moved through Christian bookstores.

Moody relaunched it with an advertising campaign in 2009, and it spent the following years on the New York Times bestseller list. The publisher now says it has sold more than 20 million copies in over 50 languages, a figure that comes from the publisher. Spin-off editions followed for children, singles, the workplace and military couples.

The book came from a Christian publisher, and critics describe its early editions as written from a conservative, traditional-marriage point of view. Chapman has said the languages apply to any couple, whatever their gender.

The five languages and the quiz that sorts them

The five categories are Words of affirmation (a compliment, a thank-you, a note), Quality time (full attention, listening, doing something together), Receiving gifts (a present for a holiday, one for no reason, something brought back from a trip), Acts of service (taking over a chore, running an errand) and Physical touch (a hug, a hand on the shoulder while pouring coffee; Chapman is clear that touch isn’t the same thing as sex).

Chapman’s model gives each person one main language and one second language; the other three matter less. People tend to show love the way they’d like to receive it, so partners with different languages can each do a lot and each feel unseen. The remedy he proposes is to learn the partner’s main language. To find it, he suggests watching three things: what the partner does most often for others, what they complain about most, and what they ask for.

The official quiz on the author’s website is free. It’s 30 pairs of statements, and in each pair the person picks the one that is “most meaningful most of the time,” even when neither fits well. The result is a main language and a second language. The full system, with a short description of each language, is on the love languages page, where a test will appear.

What researchers tested

The main check is a 2024 review by Emily Impett, Hyun Park and Amy Muise in Current Directions in Psychological Science. It went through about ten empirical studies and asked whether they support the three assumptions the book rests on: that each person has one preferred language, that there are exactly five languages, and that couples are happier when partners speak each other’s language. The authors found no convincing support for any of the three.

The book’s own example of a pair with different languages is Chapman’s marriage. He and his wife Karolyn, married since 1961, were bothered by the same thing from opposite sides: she left cabinet doors open and he closed them. By his account, once they understood the difference, their marriage moved forward. The research question is whether such differences predict how satisfied couples are, and the studies disagree about that.

One finding concerns a different factor altogether. In a 2017 study, when partners’ main languages differed, the satisfaction of both depended on how well the wife regulated her own behavior in the relationship. Self-regulation explained more of the variation in satisfaction than the match did.

The 2022 study of students’ open answers looked at behavior instead of satisfaction. Participants described a partner who texts every morning to ask whether they woke up, ate and got home safely, or who noticed cold feet and brought a blanket. Most answers fit Chapman’s categories (the blanket is an act of service, the pastries sent after a rough week are a gift). The researchers also pulled out a theme that has no category in the book: short check-in messages.

The 2022 study by Mostova, Stolarski and Matthews did find a link: couples with a larger mismatch were less satisfied. Its authors point out that this is a correlation, and that other things, such as emotional intelligence, could lie behind both. The review concludes that the evidence as a whole is insufficient. A 2006 study (Egbert and Polk) that tested a 20-item scale did find five factors in the answers; that checked the scale and didn’t test whether matching helps a relationship.

Why one language is too simple

The quiz makes people choose between two statements, and Impett argues that this forces trade-offs that don’t exist in real life. When the same five kinds of behavior are rated one at a time, people give high ratings to all five. Someone who picks quality time in the quiz may also light up when a partner praises their work in passing, and be moved when that partner fixes a broken lock. The quiz still returns a main language and a second one, even when the five scores are nearly level, so the “main language” is partly a product of the format.

Gery Karantzas, who reviewed the research in 2023, adds that no data on the reliability or validity of the official quiz has been published. Millions of people have taken it, but a count of test-takers isn’t evidence that it measures what it claims. Maha Khawaja adds that the result, used as a verdict, can end a conversation early: “we have different languages” sounds like an explanation, and the couple stops asking what each of them would actually like.

The 2025 study found something similar in people’s lives: fewer than half of its participants had a main language that stood out. What predicted relationship quality better was a wider set of behaviors, among them verbal approval, support for a partner’s personal goals and support in hard times. Chapman’s five categories leave out the support for personal goals, which may matter more to couples with egalitarian values. Impett also notes that Chapman’s own sample was mostly white, religious, heterosexual couples with traditional views.

Chapman himself doesn’t treat the number five as fixed. He’s said he was never dogmatic about it. When a man in an interview suggested chocolate as a sixth language, Chapman answered that chocolate you buy is a gift and chocolate you make is an act of service. He also says a language can change with circumstances: a mother of two small children, for instance, may suddenly care most about help.

There’s another way to look for a type. CoreTypes starts from the idea that an answer to a question about yourself shows how you see yourself, while the type itself comes through in the meaning of what you say: what you notice first and what you keep coming back to without being asked. The free CoreTypes test is built on the same idea. It has 36 questions, each with four answers: sets of words, ready-made phrases or a short scene. It takes about ten minutes and doesn’t ask you to sign up, and the result is one of 16 types with a free type page. Take the CoreTypes test.

What’s still useful

The researchers aren’t dismissive. Impett has said the book works better than the quiz, and that all five kinds of behavior Chapman describes do matter. Another researcher, Martin Graff, wrote that even if the theory isn’t science, thoughtful expressions of love have value on their own. Khawaja called the model a thin starter vocabulary for talking about care, useful as a way in and weak as a diagnosis. Karantzas points out that attention to a partner’s needs helps people feel understood, an idea that appears in many relationship models and isn’t unique to Chapman’s.

Impett and her colleagues offer their own picture: love as a balanced diet, a mix of different nutrients, instead of one language to be learned. Partners, on this view, are better off being fluent in all five behaviors than specializing in one.

The ideas translate into small things. Chapman’s newsletter once suggested kissing a partner every morning and evening for a week and noticing the response. In one of Chapman’s examples from military couples, a wife traced her hand on paper and mailed it, and a husband left behind a jacket for his wife to wear in place of a hug. Touch could travel that way.

The model also has limits that its critics point out. Someone with chronic pain or recovering from an illness may not want more touch, and “learn your partner’s language” doesn’t tell a partner to ask first. Patience and other forms of closeness can matter more in those months.

The most practical use of the model costs nothing. A couple can ask each other, once a month, what would feel good this week, and write the answers down, since the answers change with a new job, a baby or a rough winter. The five categories give them words for the conversation; for a list to start from, there’s the love languages page on Typologica, where a test will appear.