In 1938, a Harvard study began following 268 sophomores, among them a young John F. Kennedy. Nearly 90 years on, their lives have weighed in on a question Aristotle, Epicurus and Seneca pondered: do we need other people to live well?

In 1938, researchers at Harvard began a study of the university's sophomores, and over the next several years they enrolled 268 of them.

As the Harvard Gazette noted in 2017, the early researchers believed that "physical constitution, intellectual ability, and personality traits determined adult development," so they measured skulls, brow bridges and moles, and even analyzed handwriting. Among the recruits was a young John F. Kennedy.

The study was not designed to settle a philosophical argument. But nearly 90 years of these men's lives bear on one of philosophy's most enduring questions: do we need other people to live well?

In the everyday sense, the study suggests the answer is yes. The philosophers, though, were arguing about something sharper, and the study can't settle that. What it can do is make the distinction easier to see.

Aristotle and Epicurus

Aristotle, the Greek philosopher who studied under Plato in the 300s BC, opens his discussion of friendship in the Nicomachean Ethics with a strong claim. In W. D. Ross's translation, he writes that "without friends no one would choose to live, though he had all other goods."

A perhaps sharper argument comes in Book IX, where Aristotle takes up a close cousin of this article's question: whether the happy man will need friends. Some say, he notes, that a person who is supremely happy and self-sufficient has no need of friends, because he already has everything good.

Aristotle rejects this. It is strange, he writes, to make the supremely happy man “a solitary,” since “man is a political creature and one whose nature is to live with others.” His conclusion is unambiguous: “Therefore the happy man needs friends.”

For Aristotle, then, friendship is not a pleasant extra.

Epicurus taught that the good life is calm and simple. On friendship, his language is emphatic. In his Principal Doctrines, in R. D. Hicks's translation, he writes: "Of all the means which are procured by wisdom to ensure happiness throughout the whole of life, by far the most important is the acquisition of friends." The next doctrine says that "nothing enhances our security so much as friendship."

Epicurus therefore gives friendship an important place among the things that make a happy life possible. But his position is not identical to Aristotle's. For Epicurus, the ultimate good was pleasure, by which he didn't mean constant indulgence or excitement, but freedom from bodily pain and mental disturbance.

Friendship contributed to that pleasure by providing security and peace of mind. As one of the Vatican Sayings, a collection of maxims from Epicurus and his circle, puts it (Vatican Saying 34, in Cyril Bailey's translation), “It is not so much our friends' help that helps us, as it is the confidence of their help.” Knowing that friends will be there when we need them can itself bring peace of mind.

That being said, some of his statements also suggest that friends matter for more than what they can provide. Modern scholars disagree about how to reconcile this with his wider philosophy, which treats pleasure as the ultimate good.

So Aristotle and Epicurus both put friendship near the centre of a good life, but for different reasons. Aristotle explicitly argues that the happy person needs friends. Epicurus saw friendship as one of the most important sources of happiness, though it remains unclear whether he valued friends mainly for what they provide or also for their own sake.

Seneca's distinction

The Stoics, a school founded around 300 BC, are often cast as the people who said we don't need anyone. That isn't quite what they said.

In Letter 9 to Lucilius, in Richard Mott Gummere's translation, the Roman Stoic Seneca does not make friendship necessary for the wise person's happiness. But he doesn't say we should do without friends, either. His "wise" person is drawn into friendships by "natural promptings," not by "his own selfish needs." Being self-sufficient, Seneca explains, means "that he can do without friends, not that he desires to do without them."

Why can Seneca's wise person do without them? He says the wise man "is sufficient unto himself for a happy existence, but not for mere existence." To stay alive he needs many things; to live well "he needs only a sound and upright soul, one that despises Fortune."

For Seneca, friends remain valuable without being necessary for the wise person's happiness. The sage may love his friends dearly, Seneca writes, "yet all the good will be limited to his own being."

Then he draws a comparison. Epicurus, he says, wanted friends so "that there may be someone to sit by him when he is ill." Seneca turns that around: the wise person wants "someone by whose sick-bed he himself may sit."

So all three agreed that friends matter. Aristotle explicitly makes virtuous friendship part of the happy life. Epicurus assigns friendship a role in securing happiness. Seneca makes the sharper claim for self-sufficiency: friendship is deeply desirable, but the wise person's happiness does not depend on possessing it.

Put simply, it's the difference between valuing friendship and depending on it.

What the Harvard study showed

The Harvard study began with the university men. In the 1970s, it absorbed a second group: 456 men from inner-city Boston, first recruited for a separate project called the Glueck Study.

Across the two groups, relationships repeatedly emerged as one of the factors most closely associated with lives that went well.

George Vaillant, the psychiatrist who led the study from 1972 to 2004, came to see relationships as central to how the men aged, the Gazette notes.

Robert Waldinger, who later became the study's director, has described a striking relationship between midlife connection and later health. In a 2015 TED talk, quoted by the Gazette, he said: "The people who were the most satisfied in their relationships at age 50 were the healthiest at age 80."

A separate published analysis from the same research programme allows a closer look at a narrower question: how relationships and day-to-day well-being were connected.

In a 2010 paper in Psychology and Aging, Waldinger and Marc Schulz examined 47 older couples: men from the original Harvard group and their partners.

For eight days, the researchers interviewed the participants separately about their happiness, physical discomfort and time spent with other people. They also measured how satisfied each person was with their relationship.

An interesting finding concerned the quality of their marriages. People who were more satisfied with their marriages showed less of a drop in happiness on days when they experienced pain or physical limitations. Time spent with other people went with happier days in general, but it did not cushion the hard days in the same way.

This was a small study of older couples, conducted over just eight days, and it cannot establish that relationships caused the differences in happiness. Nor does it directly establish Waldinger's much longer-term claim about relationship satisfaction at 50 and health at 80. But it shows something more specific than the broad claim that relationships are good for us: the quality of a close relationship may matter for how people experience difficult days.

There is another caution worth making. Aristotle, Epicurus and Seneca were principally debating friendship, whereas the Harvard research examined a much wider set of relationships, including marriage, family, friends and community. The research therefore bears on the broader importance of human connection rather than directly testing the philosophers' accounts of friendship.

The original Harvard group was also narrow: all male, all of them students at an elite college around 1940, chosen partly because they seemed healthy and promising. And an observational study cannot by itself show that good relationships caused good health. Healthier, more emotionally stable people may also find it easier to maintain relationships.

Still, across decades of observation and in more focused investigations, the findings point in a similar direction: our relationships are closely connected with how well our lives go.

So do we need other people to live well?

Aristotle argued that even a person fortunate enough to possess everything else good would still need friends to be happy. Nearly 90 years of Harvard research point in a similar direction: relationships are closely connected with how happy and healthy people are.

There is, however, a distinction. The Harvard researchers examined measurable aspects of well-being, including reported happiness and physical health. Their findings show the importance of relationships, but they cannot establish Aristotle's stronger claim that friendship is necessary for a fully happy life.

The findings are also consistent with Epicurus's view that friendship contributes to happiness and security. They cannot, however, resolve the philosophical question of whether friendship matters chiefly because of what it provides or whether it is also valuable for its own sake.

Nor do the findings disprove Seneca. He wasn't arguing that we should do without other people: in the same letter he says we hate solitude and crave society, and that even the wise man could not live without human company. His argument concerned the Stoic ideal of a wise person who values friendship but whose happiness ultimately depends on virtue rather than on circumstances beyond their control.

All this said, nearly 90 years of research give us good reasons to take relationships seriously. They do not settle whether the perfectly wise person could remain happy without them. Aristotle might say that such a life would be missing something essential. Seneca might say that the wise person could bear the loss without losing his happiness.