Cheating and infidelity: which signs actually tell you anything

A sign turns dread into an activity. You cannot read another person's mind, but you can count how often the phone goes to the bathroom and note that they started locking it in March. A list of signs is what people make when the thing they need to know is hidden and the waiting is unbearable.

Before any checklist works, two people would have to mean the same thing by the word, and surveys keep finding they do not. Some count a single kiss; others count sexting, a secret confidant, pornography, or an old flame liked on a photo. The line sits wherever a couple drew it, or more often never discussed until it was crossed.

The most defensible numbers come from the US General Social Survey. Analyzing it, Nicholas Wolfinger found about 20% of ever-married men and 13% of ever-married women reported extramarital sex at some point. Those are lifetime figures counting only intercourse, which is why the same survey can sit next to a talk-show claim of 50% without either being a lie.

A survey that asks about sex outside marriage undercounts what most people mean by cheating: it misses the emotional affair, the secret texting, the relationship that never became physical. Widen the definition and the rate climbs; narrow it and it drops. Any single percentage is really a percentage plus a definition, usually left unstated.

The narrowest and oldest category: sexual contact with someone outside the agreed relationship. It is the version the surveys measure, the version the law once cared about, and the one nearly everyone agrees counts. It is also the easiest to define, which is why it dominates the numbers even though it is not what most breakups turn on.

Intimacy, secrecy and priority given to someone outside the relationship without physical contact: the person texted first with good and bad news, the closeness hidden from the partner. It has no legal existence and no clean test, which is why couples fight hardest here. One person experiences a betrayal the other can honestly say involved no sex at all.

The category the phone created. Sexting, dating apps kept live inside a relationship, secret accounts, an ex kept warm in the messages. The lighter end got a name, microcheating, for the small acts that feel like breaches without crossing an obvious line: a flirty comment, a hidden like, a saved number under a fake name.

The modern version is behavioral surveillance dressed as intuition, and the items barely change from list to list.

A checklist that predicts both more sex and less sex, both new fights and new distance, both guilt and accusation, has arranged itself so that almost any change confirms it. That is not a diagnostic tool. It is a shape that fits whatever you already feared, which is why it feels so accurate the week you go looking.

The usual markers are secrecy and priority: hiding the contact, and taking problems to that person before the partner. A friendship becomes an affair, in most accounts, at the point it is concealed and starts to displace the partnership, not when attraction appears. Attraction to other people is ordinary; the hiding does the damage.

The popular answer is confident and gendered: men care about the sex, women about the feelings. The research here shows how a striking finding can wobble once the method is examined.

In a 2021 study in Archives of Sexual Behavior, Dylan Selterman and colleagues surveyed roughly 500 people who had cheated and sorted their reasons into eight categories: anger at a partner, low self-esteem, love that had faded, low commitment, a wish for variety, feeling neglected, sexual desire, and simple opportunity. Sex was one motive among several, often not the main one.

A recurring finding is dull and uncomfortable: a lot of infidelity tracks opportunity. More time around potential partners, more privacy, more travel, more nights out separately, and a social circle where it happens all raise the odds. Prevention that relies only on picking a trustworthy person, ignoring the situations a relationship puts itself in, works with half the picture.

Two comfortable stories take a hit from the data. The first is that cheating always means the relationship was bad; plenty of people who report being satisfied still cheat, which is why satisfaction alone is a weak predictor. The second is that it is always about sex; the motive research surfaces anger, neglect and self-esteem as often as desire.

A suspicion built from something you actually noticed can be dated, described and checked against what happens next. A generalized fear cannot: it has no anchor, no off switch, and it often predates this relationship, attaching to whoever is nearby. What separates them is not how strong the feeling is, but whether it points at anything you could write down.

Researchers gave partner monitoring a clinical name, interpersonal electronic surveillance. Studies by Amy Muise and colleagues in 2009 and by Rachel Elphinston and Patricia Noller in 2011 found that monitoring a partner online is linked to more jealousy and lower satisfaction, not less, and feeds a self-reinforcing loop.

There is far less hard data here than the confident advice suggests, partly because the couples who split are gone before anyone can study them. Kristina Coop Gordon and Donald Baucom built a structured forgiveness program for couples recovering from an affair, and small studies reported reduced distress, but the samples were tiny. Anyone quoting a precise stay-together success rate is going past the evidence.

From the clinical work that exists, recovery is reported as likelier when the affair genuinely ended, when the partner who cheated took responsibility rather than blaming the betrayed one, when the couple could understand what the affair was about, and when both actually wanted to rebuild. What does not predict it is the intensity of the first promise to change.

Robin Dunbar argued that gossip is a kind of social grooming, and that much of ordinary conversation is talk about other people. Work by Matthew Feinberg, Robb Willer and colleagues in 2014 found some gossip is prosocial: warning others about someone untrustworthy helped groups cooperate. Talking about who cheated on whom is partly that ancient function running: a community keeping an unofficial record of who can be relied on.

What is new is the speed and the permanence. The 2015 Ashley Madison breach dumped millions of accounts into public view at once, and much of the site's female traffic later turned out to be fake. Screenshots and public call-outs now reach thousands of strangers in hours. The old gossip died in a week; the online version is searchable, one-sided by design, and hard to correct once it is wrong.

On these, one side asserts something the other has checked and found wanting. Everywhere else, the disagreement is about vocabulary rather than fact.