Sternberg's Triangular Theory of Love: The 8 Types of Love

Sternberg's own definitions are worth reading closely, because the everyday meanings of these words are looser.

Intimacy is "closeness, connectedness, and bondedness": trusting someone, telling them things, feeling understood, wanting good things for them. Passion is the drive toward "romance, physical attraction, sexual consummation", along with the longing to be near someone. Decision/commitment has two halves: in the short term, the decision that you love this person; in the long term, the commitment to keep that love going. Sternberg called intimacy the "warm" part, passion the "hot" one and commitment the "cold" one.

The three behave differently, and this is the most useful part of the theory. Passion is the least stable and the hardest to control: you can't decide to feel attracted to someone. Commitment is the easiest, because it's a decision. Passion matters most in short relationships; intimacy and commitment carry long ones.

Our Am I in Love quiz is built on these three parts. You pick one person before you start, then rate 48 statements, 16 for each part, which takes about 8 minutes. A part counts as present when your average answer reaches 3.5 on the 1 to 5 scale, so a row of neutral answers doesn't read as love. The pattern of present and missing parts names one of the eight kinds below, and that result is free. The full report shows how strong each part is. The statements are our own, written for the quiz; they are not Sternberg's published scale.

Consummate love is the full triangle: strong intimacy, strong passion and a firm commitment, all aimed at the same person. It's the answer most people would give if asked what they want from a partner, and it's the only one of the eight that has everything.

Sternberg compared it to a weight-loss target. Getting there is often easier than staying there, and people tend to notice the loss late, once it's well under way. The three parts run on different clocks: passion fades on its own unless something renews it, intimacy can go quiet without anyone noticing, and commitment slips when the relationship starts to flag. A couple can drift from consummate to companionate love without a single argument.

He also argued that it isn't always fragile: many parents' love for their children has all three parts and holds without much effort. In couples, keeping it looks unglamorous: conversations that aren't about logistics, time alone together, and new things done as a couple (there's research on that below).

Companionate love is deep friendship plus commitment, with less of the physical charge. Passionate love is the intense, longing kind: Elaine Hatfield's definition, widely used in research, calls it "a state of intense longing for union with another" (quoted in Bode and Kushnick's 2021 review, doi:10.3389/fpsyg.2021.573123).

Researchers have long described relationships as moving from the second toward the first. Susan Sprecher and Pamela Regan (1998, Sociological Inquiry, doi:10.1111/j.1475-682X.1998.tb00459.x) tested it with 197 couples at different stages, from dating to marriage. Passionate love was more sexual than companionate love and declined with time. Both kinds went with satisfaction and commitment, but companionate love was the more strongly linked to satisfaction.

Sternberg's framework adds one thing to the older two-way split: it separates romantic love (closeness plus passion) from infatuation (passion alone).

We'd push back on the idea that companionate love is the consolation prize. In Sprecher and Regan's couples it was the one more closely tied to satisfaction. The honest risk is a gap: if one partner is content with companionate love and the other misses the passion, that difference is worth saying out loud.

Sternberg drew a different curve for each part. Passion rises fast, peaks, then settles to a steadier level as the two of you get used to each other. He likened it to habituation, and noted that losing the person can drop you below where you started, which is why a breakup can feel like withdrawal. Commitment follows an S-shape: slow at first, then faster, then leveling off if the relationship lasts. Intimacy is the odd one: as partners become predictable, felt closeness can fade even while the bond underneath keeps growing. He suggested small interruptions, a trip apart or a holiday that breaks routine, show how much is really there.

Studies mostly bear this out. Robert Lemieux and Jerold Hale (2002, Psychological Reports, doi:10.2466/pr0.2002.90.3.1009) compared 446 people from casual dating to marriage: the longer the relationship, the lower the passion and intimacy and the higher the commitment. Michele Acker and Mark Davis (1992, Journal of Social and Personal Relationships, doi:10.1177/0265407592091002) found the predicted fall in passion only among women, no general fall in intimacy, and commitment as the strongest predictor of satisfaction, especially in the longest relationships. A study of 2,791 people aged 12 to 88 (Sumter, Valkenburg and Peter, 2013, doi:10.1177/0165025413492486) found adults over 50 reported less passion and intimacy than younger adults but just as much commitment.

Passion fading on average isn't passion disappearing. In a random US sample of 274 married people, 40% of those married more than 10 years said they were "very intensely in love" (O'Leary and colleagues, 2012, doi:10.1177/1948550611417015). Brain scans of 17 people married an average of 21.4 years showed responses to their partner's photo in the same reward areas seen in new love (Acevedo and colleagues, 2012, doi:10.1093/scan/nsq092). What tends to go is the obsessive edge of early love, and in long relationships that edge went with lower satisfaction (Acevedo and Aron, 2009, doi:10.1037/a0014226).

One finding is easy to use: couples who did a novel, exciting 7-minute task together reported better relationship quality afterwards than couples given a mundane one (Aron and colleagues, 2000, doi:10.1037/0022-3514.78.2.273).

Sternberg turned the theory into a questionnaire, the Triangular Love Scale, and published its validation in 1997 in the European Journal of Social Psychology. As described by Vicente Cassepp-Borges and Luiz Pasquali (2012, doi:10.1590/S0103-863X2012000100004), it has 45 statements, 15 for each part, each rated from 1 (not at all) to 9 (extremely), with a blank where you write the name of the person you have in mind. Its internal consistency was high, with each part's alpha above .90.

Since then it has been tested widely. A study across 25 countries with 7,332 participants found the scale measured the same thing in every culture tested, and that the three parts differed with relationship length in the way the theory predicts (Sorokowski and colleagues, 2021, Journal of Sex Research, doi:10.1080/00224499.2020.1787318). Shorter versions exist too, including the one used with adolescents and adults in the lifespan study above.

The scale's known weak point is that the three parts move together. In Sternberg's own data they correlated at .71 to .73, and an earlier check by Bernard Whitley (1993, doi:10.1177/0265407593103013) found the parts highly correlated and many items loading on more than one of them. We come back to that below.

Our quiz is a separate set of 48 statements we wrote, rated 1 to 5, built on the same three parts and meant for one question: which kind of love you feel for one person right now.

The part of the theory people skip is that every relationship has more than one triangle. There is the love you feel, the love you think your partner feels, and the love you'd ideally want. Each partner has their own set.

Sternberg and Michael Barnes tested this with 24 Yale couples, described in the 1986 paper. How people felt about their real partner related to satisfaction far more strongly (correlations of .55 to .75) than their picture of an ideal partner did (.28 to .41). And how you think your partner feels about you mattered about as much as how you feel about them.

This is why two people can describe the same relationship differently: one has a passion-heavy triangle, the other a commitment-heavy one, and the mismatch shows up as a vague sense that something is off. If you and your partner each take the Am I in Love quiz about the other, comparing the two results is a concrete way to start that conversation. For differences in daily life rather than feelings, our compatibility test looks at six areas where couples fit or clash.

Searches for "types of love" often lead to lists of Greek words. The best-known research version is John Alan Lee's six love styles from 1977, which Sternberg discussed in his paper: eros (passion for a physical ideal), ludus (playful, game-like love), storge (slowly growing affection and companionship), mania (obsessive, jealous love), agape (selfless love) and pragma (practical, list-based love).

Sternberg mapped them onto his triangle. Eros sits close to infatuated love, and mania is, in his phrase, "infatuated love gone berserk". Storge is close to companionate love. Ludus and pragma he didn't count as kinds of love at all: ludus is a way of playing games inside any relationship, and pragma is a way of shopping for a partner.

Lee's styles describe how a person tends to love; Sternberg's types describe one relationship at one moment. For your own habits across relationships, our love styles test looks at that side, and our guide to the 5 love languages covers a popular framework with much thinner research behind it.

We like this theory, and it has real problems.

The biggest is that the three parts are less separate than the triangle suggests. When researchers measure them, people who score high on one usually score high on the others. A meta-analysis of 81 studies and 19,387 people by James Graham (2011, doi:10.1177/0265407510389126) pooled several popular love scales and found three broader factors underneath: general love, romantic obsession and practical friendship. A 2020 study of 610 people in relationships found that the scale's answers fit best when the three parts are grouped under one general love factor (Merino and Privado, doi:10.1017/sjp.2020.3). Love may be one big thing with three faces more than three independent forces.

Second, the eight types are cut-outs. Sternberg said so himself: "Most loving relationships will fit between categories, because the various components of love are expressed along continua, not discretely." Real relationships sit between labels, which is why our full report shows how strong each part is, not only which type you landed on.

Third, almost everything rests on what people say about their own feelings at one moment. A bad week can make a companionate marriage read as empty love.

None of this makes the theory useless. It gives you three questions to ask about any relationship (how close are we, how much do we want each other, how much have we chosen this) and words for the answers. It can't predict whether a relationship will last, and neither can a quiz. For patterns in how you attach to partners in general, our attachment style test is the better tool.