LOVE LANGUAGE
QUIZ

Real as a vocabulary, shaky as a taxonomy. The five love languages give people a way to say "I want to be loved like this, not like that", which is genuinely useful. What the research does not support is the stronger set of claims underneath: that you have one primary language, that there are exactly five of them, and that matching your partner's is what makes a relationship work.

That is not a fringe view. It is the conclusion of a 2024 review in a mainstream psychology journal, and it is worth reading before you take the result of any five-question quiz too seriously.

Key Takeaways

• The model comes from The Five Love Languages: How to Express Heartfelt Commitment to Your Mate, published by Gary Chapman in 1992. It is a marriage counsellor's book, not a research programme.
• Impett, Park and Muise reviewed the evidence in 2024 and found it "does not provide strong empirical support for the book's three central assumptions".
• Those assumptions are that each person has a preferred love language, that there are five of them, and that couples are more satisfied when partners speak one another's preferred one.
• The empirical picture is not empty. Studies do find that feeling well loved predicts satisfaction. What they struggle to show is that the preferred category outperforms the others.
• The reviewers propose a different metaphor: love as "a balanced diet in which people need a full range of essential nutrients", rather than a language you learn to speak.

Where did the five love languages come from?

From one counsellor, one book and no laboratory.

Gary Chapman published The Five Love Languages: How to Express Heartfelt Commitment to Your Mate in 1992, through Northfield Publishing. The Open Library catalogue record for the first edition puts it at 204 pages. Chapman was a marriage counsellor, and the framework came out of patterns he noticed in counselling sessions, written up for couples rather than for a journal.

That origin matters for how the idea should be read. The 2024 review that examined it found "a paucity of empirical work on love languages", which is the polite way of saying the framework arrived before the evidence rather than out of it. Measurement scales came later and from other people: Hughes and Camden record that Egbert and Polk built a twenty-item Love Language Scale in 2006, and that most research on the topic has not used Chapman's own inventory.

The five categories are:

Love language	What it looks like
Words of affirmation	Saying it out loud: praise, appreciation, encouragement
Quality time	Undivided attention, with the phone away
Receiving gifts	Tokens that show you were thought about
Acts of service	Doing the thing that needed doing
Physical touch	Contact, from a hand on the shoulder upward

Nothing on that list is controversial. Every item is a real way people express affection, and most people can name which ones land hardest for them. The disagreement is entirely about the structure imposed on top.

Do the five love languages hold up in research?

Not well, on the strongest published assessment of them.

In 2024, Emily Impett, Haeyoung Gideon Park and Amy Muise reviewed the model against relationship science. Their paper ran in Current Directions in Psychological Science, and its abstract opens bluntly:

Despite the popularity of Chapman's book The 5 Love Languages, there is a paucity of empirical work on love languages, and collectively, it does not provide strong empirical support for the book's three central assumptions that (a) each person has a preferred love language, (b) there are five love languages, and (c) couples are more satisfied when partners speak one another's preferred language.

The abstract goes on to give the reasons the model spread anyway, and to propose a replacement metaphor; both are quoted further down. First take the three assumptions one at a time. The per-assumption phrasings below come from the Association for Psychological Science's own companion piece to the review, written for teaching, which sets out what the evidence showed for each.

A single preferred language. The evidence suggests "people like and appreciate all five ways of receiving love, not just one". Asked to rank them, people express a preference. Asked whether the others matter, they say yes. The same objection applies to any forced-choice typology, including the classical four temperaments traced in the personality test explainer.

Exactly five categories. "The five languages show considerable overlap, which calls into question the very existence of distinct languages." This is the assumption most directly testable, since it needs factor analysis on a validated scale. Egbert and Polk's twenty-item Love Language Scale is the instrument most of that work has used, and it arrived fourteen years after the book.

Matching predicts satisfaction. Researchers "have not found a consistent positive association between relationship satisfaction and either matching or mismatching on languages". This is the claim the whole self-help apparatus rests on, and it is the one assumption of the three where the reported result is no consistent association at all rather than a contrary finding.

The review's alternative is a metaphor rather than a new taxonomy: love is "not akin to a language one needs to learn to speak but can be more appropriately understood as a balanced diet in which people need a full range of essential nutrients to cultivate lasting love".

So what does the evidence actually show?

That being well loved predicts satisfaction, which is a smaller and less interesting claim than the one the book makes.

The empirical literature is thin but not empty, and it does not all point one way.

Evidence for matching. Mostova, Stolarski and Matthews studied 100 heterosexual couples for a 2022 paper in PLOS ONE, measuring both how people express affection and how they prefer to receive it. It is a cross-sectional self-report design, so it establishes association rather than cause, whatever the paper's title suggests. They found that "the greater the discrepancy between preferred and felt LLs [love languages], the less satisfied the participants were with their relationships", with the effect stronger for men, particularly on sexual satisfaction, where the correlations ran to minus .37 for men against minus .21 for women.

Evidence that cuts the other way. Hughes and Camden surveyed 981 people in couples for a 2020 paper in the Psi Chi Journal of Psychological Research, testing whether perceiving that a partner used your preferred love language well predicted love and relationship satisfaction. It did. This is a self-report survey too, and its inputs are perceptions rather than observed behaviour, so the same caution applies. But read their numbers across the five categories and something else appears. For relationship satisfaction the variance explained ran from .11 for gifts up to .32 for words of affirmation; for feelings of love, from .17 to .26. The authors do not frame it this way, so treat what follows as a reading of their reported figures rather than a claim they make: on those numbers every category predicted the outcome. The spread is real, and words of affirmation did roughly three times the work of gifts, so a reader could fairly say some categories matter more than others. What the figures do not show is a category that carries the outcome on its own while the rest sit at nothing, which is what a theory of one primary language would need.

The honest summary is that a partner who makes you feel loved makes you more satisfied, which nobody disputes, and that the specific machinery of five discrete categories with one primary is doing much less work than the framing implies.

Why is the idea so popular anyway?

Because it solves a real problem, even if the mechanism it proposes is wrong.

The 2024 review is unusually generous on this point, and lists three reasons the model spread. It "enables people to identify important relationship needs". It "provides an intuitive metaphor that resonates with people". And it "offers a straightforward way to improve relationships".

All three are worth taking seriously. A framework that is easy to use will beat a more accurate one that nobody reaches for, which is a general fact about advice rather than a defence of this particular advice. Before the phrase existed, a person who felt unloved despite a partner doing plenty had no compact way to describe the mismatch. "You do all the washing up but you never actually say anything" is a hard conversation to start. "My love language is words of affirmation" is an easy one. The framework's real contribution may be that it makes a difficult request sayable.

That is a defensible reason to like it and a bad reason to treat a quiz result as a fact about you. Take the label as a conversation opener, not a diagnosis, and assume the other four still count.

If you want a longer look at what short quizzes can and cannot tell you, the quizzes hub covers the scoring models. The IQ test explainer sets out why a real instrument needs a norming sample before it can claim a number, and the daily puzzles are a different case of a small, legible ritual outperforming a more effortful one.

Frequently asked questions

What are the five love languages?

Words of affirmation, quality time, receiving gifts, acts of service and physical touch. Gary Chapman set them out in his 1992 book as five ways people prefer to give and receive affection, with the argument that couples run into trouble when they express love in a category their partner does not value as highly. The categories themselves are uncontroversial descriptions of real behaviour; the contested part is the claim that they form five distinct types with one primary.

Are the five love languages scientifically valid?

Not in the sense the phrase implies. A 2024 review in Current Directions in Psychological Science concluded that the published work "does not provide strong empirical support" for the book's three central assumptions. Summarising that review for teaching, the Association for Psychological Science put the second problem plainly: the five categories "show considerable overlap, which calls into question the very existence of distinct languages". It is a popular framework with a thin evidence base, which is different from being disproved, and different again from being validated.

Does speaking your partner's love language actually help?

Sometimes, and probably not for the reason the book gives. One 2022 study of 100 couples found that a bigger gap between how someone wants to be loved and how they feel loved went with lower relationship satisfaction. A 2020 study of 981 people reported figures that, read across the five categories, show every one of them predicting satisfaction rather than only the preferred one. Both results are consistent with the simpler idea that expressed affection helps, whatever form it takes.

Can you have more than one love language?

On the evidence, most people effectively do. Summarising the 2024 review, the Association for Psychological Science reports that people "like and appreciate all five ways of receiving love, not just one", which is why the reviewers propose thinking of love as a balanced diet needing "a full range of essential nutrients" rather than a single language. A quiz that hands you one winner is compressing a ranking into a label, and the runners-up have not stopped mattering.

Who invented the five love languages?

Gary Chapman, a marriage counsellor, in The Five Love Languages: How to Express Heartfelt Commitment to Your Mate, published by Northfield Publishing in 1992. It came from his counselling practice rather than from a research programme, which explains both its readability and the shortage of empirical scaffolding underneath it. Later researchers built measurement scales for the idea, notably Egbert and Polk's twenty-item Love Language Scale in 2006; Chapman did not start from one.

Sources

• Open Library catalogue record for the 1992 first edition of The Five Love Languages: How to Express Heartfelt Commitment to Your Mate by Gary D. Chapman, Northfield Publishing, 204 pages: openlibrary.org/books/OL22613246M. Read through the Open Library editions JSON endpoint. Source of the title, subtitle, publisher, year and page count.
• Emily A. Impett, Haeyoung Gideon Park and Amy Muise, "Popular Psychology Through a Scientific Lens: Evaluating Love Languages From a Relationship Science Perspective", Current Directions in Psychological Science, volume 33, issue 2, pages 87 to 92, published online 12 January 2024 under a CC BY licence (bibliographic details confirmed against the Crossref record). The publisher's page returns a JavaScript challenge to plain fetches, so the abstract was read verbatim from the Semantic Scholar catalogue record for the paper, which carries it in full: the paper's Semantic Scholar record. Everything quoted here from the review itself comes from that abstract: the three central assumptions, the "paucity of empirical work" finding, the three reasons for the model's popularity, and the balanced-diet metaphor. The block quotation in the body is the abstract's opening, not the whole of it.
• Association for Psychological Science, "Empirical Evidence Is My Love Language", the Teaching Current Directions companion to that review: psychologicalscience.org. Source of the per-assumption findings quoted here, including "people like and appreciate all five ways of receiving love, not just one" and the overlap finding. Those three phrasings are the APS companion's, not the review's own abstract.
• Olha Mostova, Maciej Stolarski and Gerald Matthews, "I love the way you love me: Responding to partner's love language preferences boosts satisfaction in romantic heterosexual couples", PLOS ONE, 2022, open access at PubMed Central: pmc.ncbi.nlm.nih.gov/articles/PMC9216579. Source of the 100-couple sample, the discrepancy finding and the gendered correlations.
• Jennifer L. Hughes and Abigail A. Camden, "Using Chapman's Five Love Languages Theory to Predict Love and Relationship Satisfaction", Psi Chi Journal of Psychological Research, volume 25 number 3, 2020, pages 234 to 244, open access PDF: psichi.org. Source of the 981-participant sample and the variance-explained figures across all five categories.