Relationship styles

Ask what kind of partner someone is and you will almost always get an attachment answer — anxious, avoidant, secure. That framework earned its place and it has its own entry: attachment theory covers what it is, and earned secure attachment covers the part people most need, which is that it moves.

This entry is about everything else, because attachment has crowded out a set of findings that are better evidenced, more specific and considerably more useful — and that share one property the style-quiz genre tends to hide.

The differences that predict what happens to a relationship are mostly differences in what people do, not in what they are. They are habits with names, most of them are learnable, and the research on them is decades old.

Rusbult's investment model is the most durable structure in this literature, and the reason it matters is that it separates two things everybody runs together: how good this is and how likely you are to stay.

Three things predict commitment, and satisfaction is only one:

• Satisfaction — how rewarding the relationship is relative to what you expected.
• Investment — what is tied up in it that leaves with it. Years, a flat, shared friends, a language you only speak with them, a version of yourself that only exists in this.
• Quality of alternatives — how good the realistic options are, including the option of being alone.

Le and Agnew's meta-analysis, pooling decades of studies, found all three predict commitment, and commitment predicts whether people are still together later. That is the tidy part.

The untidy part is what it implies. Investment and alternatives raise commitment independently of satisfaction — which means the model explains, without any appeal to pathology, why people stay in relationships that stopped being good. There is nothing wrong with them. They have twelve years and a mortgage in it and they are forty-one. That is the mechanism, and naming it is more useful than a diagnosis; it is the relationship version of what sunk cost in careers describes, and it is why one-sided relationships are so stable.

Where people genuinely differ is in the weighting. Some people's commitment tracks satisfaction closely and moves fast when it drops. Others are anchored by investment and barely notice satisfaction until it is far gone. Neither is a virtue. They just fail differently.

Canary and Stafford asked what married people actually do to keep a relationship going, and got a list short enough to remember:

• Positivity — being pleasant to be around, on purpose, including when you do not feel like it.
• Openness — talking about the relationship itself rather than only through it.
• Assurances — saying out loud that there is a future, and that they are in it.
• Social networks — spending time with shared people, being someone's partner in public.
• Sharing tasks — doing your share of the domestic work, which is a relationship behaviour and not a logistics one.

Ogolsky and Bowers's meta-analysis found these behaviours correlated with satisfaction, commitment, love and liking across the literature. It is not a surprising result. It is worth stating anyway, because the amount of energy people put into working out their type is wildly out of proportion to the amount they put into the five things with an evidence base behind them.

Note what kind of variable this is. Maintenance is not a trait you have; it is a set of behaviours that vary week to week, drop off predictably once a relationship feels settled, and can be resumed on a Tuesday. A person who does little of it is not an avoidant type. They are a person who has stopped doing something.

The sharpest behavioural difference between people in relationships shows up in a very narrow window: immediately after the other person does something inconsiderate.

Rusbult and colleagues called this accommodation — the willingness to respond constructively when your partner has just behaved badly, rather than reciprocating. There are four moves available, and everyone uses all of them, in different proportions:

• Voice — raise it, actively, in a way aimed at fixing it.
• Loyalty — wait it out, hold position, let it pass.
• Exit — escalate, threaten the relationship, leave the room or the conversation.
• Neglect — withdraw effort, go cold, let it rot quietly.

The first two are constructive, the second two destructive, and the finding is that the destructive impulse is usually the faster one. Accommodation is what happens when something overrides it, and what overrides it is largely commitment. This is where a lot of what gets described as compatibility actually lives: not in whether two people clash, which they will, but in what each of them reaches for in the first ten minutes afterwards. Conflict and repair is the entry on the second half of that.

Neglect deserves a note, because it is the one that goes unnamed. It produces no argument and therefore no record, and a relationship can lose three years to it while both people report that they never fight.

Knee's work on implicit theories identified a difference that sits underneath all of the above: what people privately believe a relationship is.

Destiny beliefs hold that two people are either suited or they are not, and that this is a fact to be discovered. Growth beliefs hold that relationships are built, and that difficulty is part of building rather than evidence against.

Knee found these predict different behaviour, especially early. Destiny-oriented people treat early friction as diagnostic — a sign they have read the situation wrong — and are quicker to leave when it appears. Growth-oriented people treat the same friction as work and stay to do it. Neither belief is correct in all cases; a destiny reading gets someone out of a bad fit quickly, and a growth reading can keep someone in something that should end.

But it is worth noticing that the destiny belief is the one the entire compatibility-test industry is built on, including the idea that a percentage could exist. What the evidence says about that is a separate entry, and it is not what most people expect: compatibility.

Read together, the picture is this. A relationship style is a description of habits — the closeness you reach for, what you hear in what is said to you, what you do when the other person is hurting, and what you reach for when they have let you down. Those habits are stable enough to be worth naming and loose enough to be worth changing, which is the only combination in which naming them is useful.

The Relationship Style test reads that doing side rather than sorting you into a type, which is the right shape for a thing that moves. Take it as a description of current habits, not of a character.

What it cannot do: tell you whether to be with someone. There is no style that is the right one, no pairing of styles that is the recommended one, and — on the best current evidence — no useful way to compute a match between two people at all. It also cannot diagnose anything. Attachment styles are not disorders and neither is a low maintenance score.

Where a professional is the right address: if what you are calling a style difference involves fear, coercion, monitoring, or being talked out of what you remember, that is not a style, and toxic relationships is the entry, not this one. And for the ordinary case — two people who love each other and have got into a shape neither of them chose — couples therapy is the thing that exists, and it is easier work before one person has already decided.