It lifts hearts and lengthens lives. It has been hailed as the ultimate good by the greatest philosophers, promoted (at least in theory) by all the major religions and deified by revolutionaries. It even defeats the common cold. The wondrous good in question is friendship. Aristotle's highest goal for men and the third plank of the French revolution - liberty, equality, fraternity - friendship is as old as humanity and as im-portant as love or justice. But while the shelves in one part of the bookshop groan with self-help books on how to snag the perfect partner, and others (usually in the basement) are packed with economic treatises on income distribution and philosophical texts on the nature of freedom, friendship barely gets a mention. Friendship is the invisible thread running through society.
Friendship may receive miserly theoretical and political attention, but its significance in our lives is, if anything, increasing. While the claim that "friends are the new family" is an overstatement, it is certainly the case that friendships figure prominently in both the lives people actually lead and the ones to which they aspire. Television programmes such as Friends and Sex and the City portray a world in which close friendships define the contours of the participants' lives: parents and children are allowed, at best, walk-on parts. Perhaps even more than the glamour, the settings and the apartments, we envy the friendships.
Advertisers know about this. Doritos ran a series of Christmas adverts contrasting a dull, grey Christmas at home with the parents with a group of young, attractive friends in a cool apartment having a fabulous time while snacking on "friendchips". As for American Airlines, it claims to have made the skies "friendly".
The little attention friendship does receive in academia is contradictory. On the one hand, scholars such as Anthony Giddens have emphasised the growing potential of "pure relationships" - untethered by ties of class or blood - as spaces within which individuals can make their own lives. On the other, theorists such as Robert Putnam have lamented the demise of friendship as an element of "social capital". Both views contain a germ of truth, but neither is true.
Meanwhile policy-makers and politicians hardly mention friendship. In part this is because of the triumph of economistic, contractual models of the world with which friendship is entirely at odds. But it is also because friendship is seen - with more than a little justification - as a private matter. While Aristotle and many Renaissance thinkers saw politics as the expression and extension of friendship, the last thing we want is a Minister for Mates.
But the strong links between friendship and other social goods - including better health, more effective job search and higher life satisfaction - should be enough to merit the subject greater attention. Similarly, the evidence for the powerful effects of peer group on both positive and negative behaviour suggests that friendship has more influence for good or ill than any Whitehall task force.
Businesses have come to realise that social connectedness and community are important ingredients in corporate success; the strongest determinant of a person's job satisfaction is the presence of a close friend at work.
Where, though, to look for compelling analyses of friendship? One school of social science sees the emergence of "families of choice", with networks of friends supplanting blood ties. We have parents and siblings; we make friends. Giddens, in The Transformation of Intimacy, pursues this line of thought and sees friendship as the relational type most suited to a contemporary world of individuality, equality, mobility and choice. The experience of the gay community in the creation of new "families" is seen as a precursor for wider society.
In fact, blood ties remain as strong as ever. Data from the official Social Trends series shows that family is as much the first port of call for support in times of crisis as it was three decades ago. What does seem to be happening is some blurring of the lines between friends and family, what Liz Spencer and Ray Pahl call a "suffusion" of friend and family in their forthcoming Friendship: the one good thing?.
"The point is not friends or family, but friends and family," says Pahl, a professor at the Institute for Social and Economic Research at Essex University. "Familial relationships are becoming more friend-like, especially as people get older."
Putnam cites evidence from the US that friendship is "on the endangered social capital list" following a 45 per cent drop in visiting friends for dinner at home and a 60 per cent drop in the number of social picnics since the 1970s.
While Putnam says he recognises the significance not only of "machers" (the civically, institutionally engaged) but also "schmoozers" (those who operate via personal relationships), informal relationships warrant just one chapter of his Bowling Alone. And like most students of social capital, he is much more interested in parent-teacher association membership than a night in the bar. Indeed, for an academic who has railed against "data-free social science" and insists on a "double source" rule for his statements, Putnam's bias against informality is shown in the section devoted to bowling. The data he reports does indeed show a sharp decline in the proportion of bowlers who are members of a league, but there is no evidence at all that people are instead bowling alone rather than with a group of friends. Bowling Alone, then, is a title in search of a single shred of empirical evidence.
There is no evidence that I know of to suggest a decline in friendship in the UK. "Our work suggests that the social capital at the micro level is strong and powerful," says Pahl. "People are engaged in healthy personal communities of friends as well as family."
The diminution of friendship to a sub-theme of social capital, in any case, misses a broader point. One of the compelling claims of the social capital literature is that strong civic associations are related to superior economic performance - that social capital produces the financial capital. But it would be a crime to distil friendship down to a symbol "F" in a formula leading to higher GDP growth. Surely a central point of prosperity is to allow the time, space and resources to enjoy companionable relations, not the other way around.
If social scientists are blinkered about friendship, contemporary philosophers are mostly blind. (An exception is Jacques Derrida's Politics of Friendship - a tour of ancient, Renaissance and Enlightenment thinking on friendship which ultimately fails to draw its own conclusions.) The Oxford Companion to Philosophy contains no reference at all to friendship.
Part of the problem is one of definition. After all, we use the term friend to denote a dizzying array of potential relationships. I am a friend to someone I have known and loved for 25 years, but I am also a friend of Sadler's Wells. Friendship is a slippery category. This, however, is where philosophers are supposed to earn their coin. Aristotle himself did a pretty good job, dividing friendships into three types: friendships for usefulness, friendships for pleasure and friendships of virtue.
The first kind of friend is the one who will get you a job or membership of an exclusive club; the second makes you laugh. But in both cases the point of the friendship is that they provide something of separate value to you. True friendship, the third kind, is valued for and of itself. There are few numerical limits on the first two kinds - I can have a vast business network and hundreds of agreeable acquaintances - but true friendship is, by definition, a limited field: if someone has many friends, they have none.
Functional and fun friendships may be short-lived. But virtuous friendship is long-term and committed and brings the greatest psychological benefits. There is plentiful research evidence showing that having at least one close friend is associated with a range of health benefits, from recovery times from cardiac illness and lowered incidence of mental illness through to greater resistance to the common cold. Technology is helping people to maintain friendships over time and distance. As Pahl says: "Close friends are often not close at all now, geographically speaking."
But friendship is not always an unalloyed good, the benefits of friendship are unevenly spread and the impact of friendship on traditional objectives of the centre left, such as equality, diversity and mobility, is mixed.
The first problem is that men are worse at friendship than women. Montaigne said of women and friendship that "their soul does not seem firm enough to endure the strain of so tight and durable a knot". Now it is widely acknowledged that women do more of the "social" work than men and have better-developed friendships skills. The feminisation of friendship has left men at a disadvantage.
Second, friendship has social and political downsides for a government committed to social inclusion: it is, by definition, exclusive. People make friends with those similar to themselves - PLUs (People Like Us). As Aristotle first pointed out, birds of a feather stick together. Nick Emler, at the University of Surrey, shows that people of a particular social class or educational background are highly likely to form friendships, or romantic relationships, with people of the same background.
Given that friends help each other, the danger is that the friendships of the affluent and successful in effect "hoard" social advantage, to the detriment of social mobility. Historically, the leisure time to cultivate friendship has been more easily available to the better-off. It is no coincidence that the word "chum" has its roots in Oxford University slang, while "crony" derives from Cambridge University. (In this sense most of "Tony's cronies" are really chums.)
Third, there is an important distinction between friendship and friendliness. Augustine, who said that health and friendship were the twin essentials of life, urged that we extend the hand of friendship to all. It is not possible to be friends with lots of people: Augustine's point was that we should treat all people as if they could be, or could become, friends. This is the definition of friendliness rather than friendship. A society of friends could have no concerns outside its own networks while a friendly society is evident from the way it treats strangers. Antisocial behaviour, road rage and intolerance towards immigrants are all evidence that, even if we have friends, we are if anything becoming less friendly.
What political inferences can be drawn? Pahl reckons the best politicians can aim for is "not to make things worse" for friendship. Certainly we do not want a national befriending scheme. But institutions, including the state, can act to improve as well as worsen the conditions for friendship.
The most obvious example is the need for a politics of time. Friendship needs time to flourish - Aristotle reckoned one and a half bushels of salt needed to be consumed together (presum- ably with other foods) before a friendship became solid. At the moment the "work-life balance" movement is focused on families. Yet parents, including fathers, spend more time with their children than ever before and at least as much time with their partners. If anything is lost in the "time squeeze" it is more likely to be friendship.
But bans on long working hours could well be counter-productive: after all, a third of us make most of our friends through work. The aim has to be to attack simultaneously cultures that force people to stay in the workplace against their will, support those who want to work and reduce the relentless drive for commercial gain that squeezes conviviality out of the office.
There is also a case for encouraging spaces in which people from different backgrounds meet and interact in order to increase the chances of cross-class friendships. Given the increase in geographical inequality, with rich people increasingly living in neighbourhoods of rich people, only NHS wards and churches are places of genuine social mixing. With the rise in private medicine and secularism, these melting pots are shrinking.
Finally, there is a friendship case for greater equality. Friendships across big social distances are rare, leading to a vicious circle of social and economic immobility. Emler cites studies showing that only 4 per cent of teenagers make friends with or date someone more than one social class away: a precursor of adult social separation. If the government is serious about social mobility - and there are those who see it as the big theme for the third term - one way to increase the circulation of our elites is to shorten the social distances across which mobility-enhancing and inequality-reducing friendships have to reach.
Friendship is a virtue with some of the appearances of a social vice. It can promote or demote social mobility; underpin tolerance or bolster discrimination; erode or sustain hierarchies. Pahl argues, following E M Forster, that a central challenge to friendship is the notion of betrayal. Who do we collectively want an individual to betray - their friend, or some wider good, such as their country? "If someone is great friends with someone at their mosque and the police - whom they have reason to distrust - is asking questions about them in relation to some terrorist activity, what do we want them to do?" Pahl asks. "In this sense MI5's job is to undermine social capital, to promote the betrayal of friends."
But there is an even deeper issue here. It may be that society could be composed of strong friendships between people of identical social backgrounds who treat everyone else with contempt, intolerance or fear. The true test of the friendliness of a community is not simply the way its citizens treat their own friends, but whether they befriend the broader social world. We need not only the care of friends, but the kindness of strangers.