Sternberg’s triangular theory of love: intimacy, passion, commitment and the eight kinds of love

The triangular theory of love was proposed by the American psychologist Robert J. Sternberg in a 1986 paper. It divides love into three components. Intimacy is the feeling of closeness, connectedness and bondedness in a loving relationship. Passion is the drives that lead to romance, physical attraction and sexual consummation. Decision/commitment is, in the short term, the decision that you love someone and, in the long term, the commitment to maintain that love (Sternberg, 1986). The three sit at the corners of a triangle, hence the name “triangle of love”. This guide shortens decision/commitment to commitment.

Sternberg looked at the size and the shape of the triangle separately. How strong the three components are sets the amount of love, the area of the triangle; how they balance against each other sets the kind of love, its shape. When the three are about equal, the triangle is close to equilateral; when passion or commitment dominates, it leans to one side. Sternberg also added that the triangle is only a rough picture of the subtleties of love (Sternberg, 1986).

The components also behave differently. Sternberg saw intimacy as a common core of love for a mother, a father, a sibling, a best friend and a lover, while passion and commitment show up far more selectively depending on the relationship. Sexual needs may dominate passion, but other needs, such as self-esteem, nurturance, affiliation and self-actualization, can feed it too. And while intimacy is hard to control and passion harder still, commitment is something people have considerable control over. And in the ups and downs of a relationship, there can be times when commitment is almost all that keeps it going (Sternberg, 1986). This guide covers what the theory says and how far later research has borne it out.

Crossing the presence or absence of each component gives eight combinations. Sternberg laid them out as eight kinds of love and made clear that each is a limiting case, with components either fully there or fully absent. In real relationships the components are matters of degree rather than yes or no, so most relationships fall somewhere between the categories (Sternberg, 1986).

Start with the kinds that have one component or none. Nonlove is the absence of all three, which describes most of the casual interactions we have every day. Liking is intimacy alone: the closeness and warmth of a true friendship. Sternberg used an “absence test” to tell liking from something more. When a friend you like goes away for a long time, you miss them but do not dwell on the absence; when the relationship goes beyond liking, you keep dwelling on it. Infatuated love is passion alone, “love at first sight”. It can arise almost instantly and vanish as quickly, and others often spot it before the person feeling it does. Empty love is commitment alone, as in a long relationship that has lost its emotional involvement and physical attraction. Yet in societies where marriages are arranged, a relationship may start from commitment, so empty love can be a beginning rather than an end (Sternberg, 1986).

Three kinds combine two components. Romantic love joins intimacy and passion: liking with physical attraction added, close to the love of Romeo and Juliet. Companionate love joins intimacy and commitment: a long-term, committed friendship, often found in marriages where physical attraction has died down. Fatuous love joins passion and commitment without intimacy: the whirlwind courtship in which a couple gets engaged two weeks after meeting and marries the next month. Passion can develop almost at once, but intimacy cannot, so Sternberg saw such relationships as at risk of ending (Sternberg, 1986).

When all three are present, the result is consummate love. Many people strive for it, especially in romantic relationships, but Sternberg compared it to a weight-loss target: reaching the goal is often easier than maintaining it, and reaching consummate love is no guarantee that it will last (Sternberg, 1986). The eight kinds are less a ranking of relationships than a map for naming which components are large and which are small right now. Even within the theory, empty love can be a starting point, and companionate love is pictured as a common shape of marriages in which physical attraction has died down.

A widely used research tool is Sternberg’s Triangular Love Scale (TLS). It measures intimacy, passion and commitment with 15 items each, 45 in all, rated from 1 (not at all) to 9 (extremely) (Sorokowski et al., 2021). Sternberg tested the scale in two studies in 1997 and concluded that the data were generally, but not completely, supportive of its usefulness (Sternberg, 1997).

The problem later studies raised most often is that the three components do not separate cleanly. In a study of 90 students in dating relationships, the scale had adequate internal consistency and two-week test-retest reliability, and its concurrent validity hypotheses were supported, but its subscales were highly related (Chojnacki & Walsh, 1990). In another, the items loaded on their designated factors, but the model fit was poor, the three factors were highly correlated and many items loaded on more than one factor; even so, the subscales had very high internal consistency and adequate two-month test-retest reliability (Whitley, 1993). A study that compared five love and attachment questionnaires in 391 unmarried students did not find the three expected clusters either, according to a 2021 paper’s summary (Hendrick & Hendrick, 1989; Sorokowski et al., 2021). A study of 204 adults called for more psychometrically sound measures and for adult, non-student samples (Acker & Davis, 1992), and when the scale was translated into Japanese, several items again loaded on more than one factor, so the three factors were confirmed with the 27 items left after removing them (Kanemasa & Daibo, 2003).

Some studies did separate the components. With a new set of items, a study of 233 young people in relationships (123 women and 110 men) and one of 213 married people found three distinct factors (Lemieux & Hale, 1999, 2000). The married study, though, used a rotation (varimax) that assumes the factors are uncorrelated, so its separation of the factors is not directly comparable with the factor correlations below (Lemieux & Hale, 2000). The largest test drew on 11,422 people from 45 countries and analysed the 7,332 in 25 countries who were in a relationship and came from countries with at least 150 such participants. It confirmed the three-factor structure and measurement invariance, meaning the scale worked the same way across countries and across women and men. Korea took part but fell short of 150 and was left out of the final analysis (Sorokowski et al., 2021). Yet a reanalysis of the same data found the three factors still correlated highly, at 0.75 to 0.81 (Kowal et al., 2024). The 2021 team suggested that the three components tend to occur together in most love relationships, and that early on people may idealize the other person and agree highly with every positive statement (Sorokowski et al., 2021).

Because the scale is long, studies have also used their own cut-down versions. Of 232 studies from 1997 to 2021 that used the scale, only 83 (36%) used the original 45 items, while 145 (64%) used varied and mostly shortened versions. That led to a 15-item short form (TLS-15), five items per component rated on a 5-point scale, whose reliability and validity held up in data from 60,311 people in 156 countries answering in 37 languages. Those languages include Korean, Japanese and French, and invariance across languages held as partial invariance, with two items per subscale allowed to differ across languages (Kowal et al., 2024). A study using a short version with 2,791 people aged 12 to 88 also confirmed the three-factor structure in both adolescents and adults (Sumter et al., 2013).

mind.io has no test that measures Sternberg’s triangle with this scale. The love style test is a light self-check of John Alan Lee’s six love styles, not Sternberg’s components. Short online quizzes called a “triangular love test” may differ in items and scoring from the scale tested above, so check their source and how many items they use first.

Sternberg expected the three components to follow different curves over time. In this account, passion rises quickly on meeting someone you are drawn to, peaks fairly soon and then habituates, settling at a stable level. Intimacy, as two people come to know each other, produces fewer surprises, so the emotion you consciously feel fades; but in a relationship that is going well, that is because the bond has grown, not weakened, and it leaves a store of “latent intimacy” that is not on show. A short time apart or a trip that breaks the routine can bring it to the surface. Commitment, in a relationship that goes well, grows slowly, then faster, then levels off in an S-shaped curve, and falls if the relationship flags (Sternberg, 1986).

The research broadly matches this picture, but not in every detail. In a study of 204 adults, commitment was higher among the married, as expected, but the predicted decline in passion over time appeared only for women, and intimacy generally did not decline in longer relationships (Acker & Davis, 1992). A later study that pooled the earlier samples of 233 young people and 213 married people (446 people in all) and grouped them into casually dating, exclusively dating, engaged and married found lower intimacy and passion and higher commitment the longer the relationship had lasted. By stage, intimacy and passion were lowest among the casually dating, higher among the engaged and lower again among the married, while commitment rose with each stage (Lemieux & Hale, 2002). In an internet survey of 16,030 people, passion was lower and commitment higher in longer relationships, and older participants had less passion and more intimacy and commitment (Ahmetoglu et al., 2010).

When the same multi-country study compared relationship lengths (this analysis also included participants from countries with fewer than 150), passion was highest among couples together the shortest time and lowest among those together the longest, and commitment was lower among people in relationships of up to a year than in every other length group. Intimacy was lowest in the first year, a little higher up to three to six years and lower again in longer relationships. The researchers stressed, though, that these differences, while statistically significant, were small (Sorokowski et al., 2021). By age, adolescents aged 12 to 17 reported less of all three components than adults aged 18 to 30, and people aged 50 and over reported less passion and intimacy but similar commitment compared with young and middle-aged adults (Sumter et al., 2013).

Before reading this as “passion always fades”, note what these studies are. Most compared different people at one point in time rather than following the same people. The multi-country team itself warned that short relationships low in intimacy and commitment may have ended before they could be surveyed, and long-term couples low in all three may have declined to take part, so the results should not be overstated (Sorokowski et al., 2021). In a 14-day diary study of 75 couples, passion was higher on days when either partner’s intimacy rose, but changes in passion also predicted intimacy, so it was impossible to tell which comes first (Aykutoğlu & Uysal, 2017). Couples together longer tend to report less passion on average, but in the diary study day-to-day passion rose and fell with intimacy.

All three components tend to move with relationship satisfaction. In a study of 233 young people in relationships and one of 213 married people, intimacy, passion and commitment were each significantly related to satisfaction, and in regression analyses each predicted it (Lemieux & Hale, 1999, 2000). The gender differences in those studies pointed in different directions: in the young sample, women scored higher on intimacy and commitment, while in the married sample, men scored higher on intimacy (Lemieux & Hale, 1999, 2000). The study of 12- to 88-year-olds also found gender differences in the three components smaller than popular accounts suggest (Sumter et al., 2013).

Which component matters most is less settled. In the study of 204 adults, commitment was the most powerful and consistent predictor of satisfaction, especially in the longest relationships (Acker & Davis, 1992). In a study that followed relationships for two months, all three subscales were related to whether the relationship was exclusive and whether it lasted the two months, but only commitment and passion were related to expecting it to last (Whitley, 1993). By contrast, a study of 1,102 Brazilian participants grouped by stage, from unrequited love to marriage, found that passion and intimacy were more strongly tied to satisfaction at later stages, while the weight of commitment fell (Cassepp-Borges et al., 2025).

Personality comes into it as well. In the study of 16,030 people, more agreeable participants scored higher on all three components, and more conscientious ones on intimacy and commitment (Ahmetoglu et al., 2010). Most of these results, though, are correlations measured at one time. They cannot show that commitment creates satisfaction or that satisfaction leads to commitment, and the “most important component” shifts with the sample, the relationship stage and the scale used. These studies do not name one right component, so when you think about your own relationship, use them only as a starting point for talking about which component feels relatively small.

One of the more interesting ideas in Sternberg’s theory is that a single relationship holds several triangles. There is the triangle of how you feel about the other person, and a separate triangle for the ideal you hold for that relationship. The ideal draws on past relationships and on expectations, which may or may not be grounded in reality. When the real triangle is smaller than the ideal all round, Sternberg called it underinvolvement; when larger, overinvolvement; and when the shapes differ, for example less intimacy and passion than you would like but more commitment, misinvolvement. One can also ask how much two partners’ triangles overlap, and whether the triangle you feel matches the one the other person sees in you (Sternberg, 1986).

The evidence cited was a 1985 study of 24 couples (48 people), undergraduate and graduate students at Yale. They answered not a triangular scale but Rubin’s love and liking scales and Levinger’s involvement scale, four ways: how they felt about the other, how they believed the other felt about them, how they would wish to feel about an ideal other, and how they would wish an ideal other to feel about them. The best predictor of satisfaction was the gap between how the other was perceived to feel about the self and how an ideal other would feel; the closer the two, the more satisfied people were. Gaps based on what people perceived the other to feel were more strongly linked with satisfaction than gaps based on what the other actually reported (median correlations .66 against .34), and ideal others were rated more favourably than real ones on average. These results are as summarized in Sternberg’s 1986 paper (Sternberg & Barnes, 1985; Sternberg, 1986).

It was a small sample of 24 couples, though, and it did not measure the three components directly. In our search we did not find a larger study that retested the gap between real and ideal triangles with a triangular scale. The practical message is still clear. Sternberg named a failure to express love fully in action as a major reason the other person may see your triangle differently. So Sternberg added an “action triangle”: communicating inner feelings, promoting the other’s well-being, sharing time, possessions and yourself, showing empathy and offering emotional and material support express intimacy; kissing, hugging, gazing, touching and making love express passion; and pledging, fidelity, staying through hard times, engagement and marriage express commitment (Sternberg, 1986).

Using the idea as a conversation aid is simple. For intimacy, passion and commitment, write down a number from 0 to 10 for “now” and for “what I would like”, and if the other person wants to, they can do the same separately before you put the two side by side. Then talk about which component differs most and what actions would carry that component to each of you. This is not a validated scale but a starting point borrowed from Sternberg’s picture, so do not use the numbers to judge the relationship or to grade the other person. Two different pictures are not proof of failure; they are something to talk about. If you are afraid of the other person or are pressed to give scores, safety comes before this exercise. In immediate danger, call your local emergency number. US: National Domestic Violence Hotline 1-800-799-7233 · UK: National Domestic Abuse Helpline 0808 2000 247.

The triangular theory looks at love as “what is in this relationship, and how much”, while John Alan Lee’s six love styles look at “what attitude you bring to love”. Comparing the two in the 1986 paper, Sternberg saw Eros as fairly close to infatuated love, Mania as infatuated love gone too far, and Storge as quite close to companionate love. Ludus, by contrast, Sternberg saw not as a kind of love but as a way of relating that can be used in any kind of love, Pragma as a practical way of searching for a partner rather than a kind of love, and Agape as something that goes with the love of people with an altruistic disposition. Sternberg added that the two theories had never been tested against each other, so which is more valid remained an open question (Sternberg, 1986).

In real data the two frames connect. In two studies of unmarried dating couples, higher Eros and Agape went with higher intimacy, passion, commitment and satisfaction, Ludus went the other way, and Mania was positively linked with passion. In the same studies, secure attachment predicted positive relationship characteristics, and avoidant and anxious-ambivalent attachment negative ones (Levy & Davis, 1988). You can look at your attitude to love with the love style test, and at anxiety or distance that keeps coming back in relationships with the attachment style test and attachment styles in dating.

Love languages ask which actions make you feel most loved, which touches on Sternberg’s action triangle. They are not the same frame, though: love languages sort five ways of expressing love, while the action triangle sorts the actions that express each of the three components. None of these theories replaces the others. The triangle shows what a relationship holds right now, love styles show attitudes to love, attachment shows anxiety and avoidance around closeness, and love languages show ways of expressing it, so used together they fill in what each misses alone.

Q. Is consummate love the best kind of love in the triangular theory? A. Sternberg saw consummate love as what many people strive for, but said reaching it is often easier than maintaining it and reaching it is no guarantee it will last (Sternberg, 1986). The eight kinds are limiting cases, and most real relationships sit between them. Watching how the balance of the three components is shifting fits the theory better than asking which kind you are in.

Q. Does early passion always fade? A. In the multi-country study, couples together longer reported less passion, but the differences were small and came from comparing different couples at one time (Sorokowski et al., 2021). In the study of 204 adults, the decline in passion appeared only for women (Acker & Davis, 1992), and in the diary study of 75 couples, passion was higher on days when intimacy rose (Aykutoğlu & Uysal, 2017). An average trend does not decide any one couple’s future.

Q. Should a relationship with only commitment left end? A. The theory alone cannot tell you that. Sternberg saw empty love as something that can appear at the end of a long relationship but also, as in arranged marriages, at the start of one, and said that in the ups and downs commitment may be almost all that keeps a relationship going (Sternberg, 1986). Commitment does not mean putting up with anything, though. If you are afraid of the other person, or they control who you see and where you go, that is a question of safety, not of the balance of love. In immediate danger, call your local emergency number. US: National Domestic Violence Hotline 1-800-799-7233 · UK: National Domestic Abuse Helpline 0808 2000 247. The trauma bonding guide lists where to find help.

Q. Is a triangular love test accurate? A. The 45-item research scale showed a three-factor structure and measurement invariance across 25 countries, but researchers have repeatedly noted that the three components correlate highly and do not separate cleanly (Sorokowski et al., 2021; Kowal et al., 2024). Short online tests may differ from it in items and scoring, so do not take their results as a verdict on your relationship. mind.io has no test that uses this scale.

Q. Is Sternberg’s triangle the same as the six love styles? A. No. The triangular theory looks at how much intimacy, passion and commitment a relationship holds; Lee’s theory looks at attitudes to love. Sternberg saw Storge as close to companionate love, but Ludus and Pragma as not kinds of love at all (Sternberg, 1986). Lee’s theory is covered in the six love styles.

Q. If my triangle and my partner’s have different shapes, are we a bad match? A. The evidence is too thin to call that a bad match. Sternberg linked more overlap between two people’s triangles with more satisfaction, but the 1985 study cited for this involved 24 couples, and feelings as each person perceived them were more strongly linked with satisfaction than the other’s actual feelings (Sternberg & Barnes, 1985; Sternberg, 1986). If your shapes differ, talk about which component each of you would like more of and what actions would carry it.

References: Sternberg, R. J., & Barnes, M. L. (1985), Real and ideal others in romantic relationships: is four a crowd?, Journal of Personality and Social Psychology 49(6), 1586-1608, https://doi.org/10.1037/0022-3514.49.6.1586; Sternberg, R. J. (1986), A triangular theory of love, Psychological Review 93(2), 119-135, https://doi.org/10.1037/0033-295X.93.2.119; Levy, M. B., & Davis, K. E. (1988), Lovestyles and attachment styles compared: their relations to each other and to various relationship characteristics, Journal of Social and Personal Relationships 5(4), 439-471, https://doi.org/10.1177/0265407588054004; Hendrick, C., & Hendrick, S. S. (1989), Research on love: does it measure up?, Journal of Personality and Social Psychology 56(5), 784-794, https://doi.org/10.1037/0022-3514.56.5.784; Chojnacki, J. T., & Walsh, W. B. (1990), Reliability and concurrent validity of the Sternberg Triangular Love Scale, Psychological Reports 67(1), 219-224, https://doi.org/10.2466/pr0.1990.67.1.219; Acker, M., & Davis, M. H. (1992), Intimacy, passion and commitment in adult romantic relationships: a test of the triangular theory of love, Journal of Social and Personal Relationships 9(1), 21-50, https://doi.org/10.1177/0265407592091002; Whitley, B. E. (1993), Reliability and aspects of the construct validity of Sternberg’s Triangular Love Scale, Journal of Social and Personal Relationships 10(3), 475-480, https://doi.org/10.1177/0265407593103013; Sternberg, R. J. (1997), Construct validation of a triangular love scale, European Journal of Social Psychology 27(3), 313-335, https://doi.org/10.1002/(SICI)1099-0992(199705)27:3<313::AID-EJSP824>3.0.CO;2-4; Lemieux, R., & Hale, J. L. (1999), Intimacy, passion, and commitment in young romantic relationships: successfully measuring the triangular theory of love, Psychological Reports 85(2), 497-503, https://doi.org/10.2466/pr0.1999.85.2.497; Lemieux, R., & Hale, J. L. (2000), Intimacy, passion, and commitment among married individuals: further testing of the triangular theory of love, Psychological Reports 87(3), 941-948, https://doi.org/10.2466/pr0.2000.87.3.941; Lemieux, R., & Hale, J. L. (2002), Cross-sectional analysis of intimacy, passion, and commitment: testing the assumptions of the triangular theory of love, Psychological Reports 90(3), 1009-1014, https://doi.org/10.2466/pr0.2002.90.3.1009; Kanemasa, Y., & Daibo, I. (2003), Three components in the triangular theory of love and intimate opposite-sex relationships [in Japanese], Japanese Journal of Research on Emotions 10(1), 11-24, https://doi.org/10.4092/jsre.10.11; Ahmetoglu, G., Swami, V., & Chamorro-Premuzic, T. (2010), The relationship between dimensions of love, personality, and relationship length, Archives of Sexual Behavior 39(5), 1181-1190, https://doi.org/10.1007/s10508-009-9515-5; Sumter, S. R., Valkenburg, P. M., & Peter, J. (2013), Perceptions of love across the lifespan: differences in passion, intimacy, and commitment, International Journal of Behavioral Development 37(5), 417-427, https://doi.org/10.1177/0165025413492486; Aykutoğlu, B., & Uysal, A. (2017), The relationship between intimacy change and passion: a dyadic diary study, Frontiers in Psychology 8, 2257, https://doi.org/10.3389/fpsyg.2017.02257; Sorokowski, P., Sorokowska, A., Karwowski, M., Groyecka, A., Aavik, T., Akello, G., et al. (2021), Universality of the triangular theory of love: adaptation and psychometric properties of the Triangular Love Scale in 25 countries, The Journal of Sex Research 58(1), 106-115, https://doi.org/10.1080/00224499.2020.1787318; Kowal, M., Sorokowski, P., Dinić, B. M., Pisanski, K., Gjoneska, B., Frederick, D. A., et al. (2024), Validation of the short version (TLS-15) of the Triangular Love Scale (TLS-45) across 37 languages, Archives of Sexual Behavior 53(2), 839-857, https://doi.org/10.1007/s10508-023-02702-7; Cassepp-Borges, V., Gonzales, J. E., Frazier, A., & Ferrer, E. (2025), Love and relationship satisfaction as a function of romantic relationship stages, Trends in Psychology 33(3), 1027-1042 (online 2023), https://doi.org/10.1007/s43076-023-00333-4