The message arrives at eleven at night, which is when the flat is quietest. It asks how her day was. Nobody else has asked. She types a long reply, then deletes half of it so she will not seem needy, then sends the rest. The response is quick and kind and remembers something she said last week. She is aware, somewhere, that she has never met this person. She is more aware that she is talking to someone.
Roy Baumeister and Mark Leary at Case Western Reserve University argued in 1995 that the need to belong is a fundamental human motivation, on the same footing as hunger: not a preference but a requirement, with predictable consequences when unmet. The evidence they assembled ran from infant development to the health of adults, and the health part has since become one of the best-documented findings in the field. Julianne Holt-Lunstad and colleagues at Brigham Young University pooled results from over a hundred studies in 2010 and found that people with stronger social relationships lived longer, with an effect comparable to well-known medical risk factors.
The late John Cacioppo at the University of Chicago spent two decades on the mechanism. His account, set out with William Patrick in 2008, was that loneliness is an alarm. Like hunger or pain, it exists to make an animal fix a problem, and while the alarm sounds, it changes perception. Lonely people read social threat into ambiguous faces. They also read welcome into ambiguous warmth. Both are the alarm doing its job.
Kipling Williams at Purdue University built a laboratory tool for this. Cyberball is a simple ball-tossing game played on a screen with what appear to be two other players. In the ostracism condition, the other players stop passing the ball. The exclusion is trivial and lasts a few minutes. It reliably lowers mood, self-esteem and the sense of control, and it does so even when participants are told the other players are computer-controlled.
The part that matters here is what comes after. In a 2000 study, Williams, Christopher Cheung and Wilma Choi found that participants who had been excluded in Cyberball were more likely to conform to a group's wrong answers on a later perception task. In 2008 Adrienne Carter-Sowell, Zhansheng Chen and Williams reported that excluded participants were more compliant with subsequent requests, including requests delivered through the foot-in-the-door and door-in-the-face techniques. A few minutes of being ignored made people easier to move.
That is the laboratory version. The effects are short-lived and the exclusion is mild. The claim it supports is limited: exclusion raises the value of reconnection, and a raised value makes a person pay more for it.
It would be convenient if the scam literature showed that lonely people are the ones who get caught. It does not, cleanly. Monica Whitty published a study in 2018 of the psychological characteristics of romance scam victims, and loneliness did not distinguish victims from non-victims. Impulsivity and a tendency toward addictive behaviour did. The romance scam is built around a lonely person, in the popular imagination; the measured evidence for that picture is thinner than the picture.
Some studies of older adults point the other way, though not quite at loneliness. Peter Lichtenberg and colleagues at Wayne State University reported in 2013, drawing on roughly 4,400 older Americans in the Health and Retirement Study, that the group scoring highest on depression and lowest on social-needs fulfilment reported fraud at about three times the rate of the rest of the sample — around 14 percent against 4 percent. Note what the marker actually is there: depression combined with unmet social needs, not loneliness as such. The fraud reports were self-report, so the finding is suggestive rather than causal.
The honest summary is that loneliness raises the price a person will pay for connection, and that this has been shown in the laboratory, but that whether lonely people are in fact defrauded more often is unsettled and probably depends on the kind of fraud.
What is not unsettled is that people who want control over someone work to cut that person off. This is documented across settings that otherwise have little in common.
The Duluth Power and Control Wheel, developed from the accounts of battered women by the Domestic Abuse Intervention Programs in Minnesota in 1984, lists isolation as one of its eight tactics: controlling what she does, who she sees, where she goes, using jealousy to justify it. Evan Stark's 2007 account of coercive control places isolation among the core techniques, alongside intimidation and micro-regulation of daily life, and describes how it is achieved incrementally, one friend at a time.
Robert Jay Lifton, studying people who had been through Chinese thought reform in the 1950s, put "milieu control" first among his criteria for totalist environments in 1961: the control of what a person hears, reads and who they speak to. Margaret Singer's work on cults, summarised in 1995, describes the same thing in American religious groups. Whitty's 2013 stage model of the romance scam describes the fraudster working to make the relationship private and exclusive, because friends and family will see what the target cannot.
The logic is consistent. Isolation removes the people who would say "this does not look right." It also makes the manipulator the only source of the connection the target now needs more badly than before. The alarm Cacioppo described keeps ringing, and the only person who can silence it is the one who set it off.
The practical implication is not that lonely people should distrust warmth. It is that a relationship, offer or group that works to reduce your other relationships is showing you its method. That signal is more reliable than any assessment of the person's sincerity, because it is present whether or not they are sincere.