Friendship may well be the most important relationship of the 21st century. As people spend more and more years of their adult lives single (and sometimes their entire lives), as family size shrinks, and as more and more people have no children at all, the place of friends in our lives grows in significance. Friendship is also important because the values at its core, such as equality, choice, flexibility, fluidity, and self-expression, are the values we often live by in 21st century Western societies.
Even so, friendship is sometimes treated dismissively. Responding to a blog post I had written about how single people have more friends than married people, and do more to maintain their ties to their friends, a reader emailed me and said that the findings “make sense.” Single people, he said, are compensating for not having more significant relationships.
Unfortunately, reputable researchers who should know better have sometimes made a similar argument. I found examples when I did the research for the very first scholarly article I ever published about single people, “Singles in society and in science,” with Wendy Morris. Some attachment researchers were claiming that only romantic partners qualify as adult attachment relationships, and even close friendships “cannot compensate.”
The scholars also waxed poetic about romantic relationships in silly ways that Wendy and I gently mocked in our article. What still makes me shake my head in dismay all these years later is that serious academics, who undoubtedly see themselves as open-minded, can be so dismissive of single people and the important people in their lives.
If it were true, though, that genuine, full-blown attachment relationships only develop with romantic partners and not with close friends or relatives, then I would just have to deal with that. But that turns out not to be true. Research has shown that single people often have relationships with friends (and family) that meet all the criteria for a true attachment relationship.
Here’s the relevant section from the 2005 article, “Singles in society and in science”:
Most early research focused on infants’ attachment to their mother. Bowlby (1979) maintained from the outset, however, that bonds of attachments were important over the entire life-span. In the last decade or so, the study of adult attachments has gained tremendous momentum. The premise has been that adults might also enjoy feelings of emotional security and protection if they have a secure attachment to another adult. They might also have more positive experiences at the form of adult exploration known as work (Hazan & Shaver, 1990). If attachment bonds are of such potential significance to adults, then it is important to know what sorts of adult bonds would qualify.
In her 1991 theoretical statement about attachment bonds across the lifespan, Mary Ainsworth described an attachment bond as one kind of affectional bond. She defined an affectional bond as “a relatively long-enduring tie in which the partner is important as a unique individual, interchangeable with none other…there is a need to maintain proximity, distress upon inexplicable separation, pleasure or joy upon reunion, and grief at loss.” Attachment bonds are special in that they meet one other criterion: the attachment figure is a source of security and comfort.
Ainsworth (1991) believed that sexual partnerships, whether same-sex or heterosexual, could become attachment bonds. However, she did not construe the sexual component as a necessary feature of an adult attachment bond. She believed that other close and enduring relationships, such as close friendships and relationships with siblings, could also become attachment bonds.
Within a few years, a set of questions had been developed to identify the people to whom adults turned for emotional security, and in 1997, Fraley and Davis reported that both romantic partners and best friends sometimes fit that role. By 1999, however, Hazan and Zeifman, despite citing the Fraley and Davis study, concluded that only sexual pair bonds could qualify as adult attachments. They made the argument that “pair bonds and infant-caregiver relationships show conspicuous similarities in the nature of physical contact, and these differentiate them from other classes of social relationships” (p. 341). The construal of parent-child touching as fundamentally similar to the touching between sexual partners could, we think, be reasonably regarded as startling.
There was more: “The social provisions of pair bonds are sufficiently distinctive that other social relationships–even close friendships or kin ties–cannot compensate for their loss” (p. 344). Perhaps. But they do not consider the complementary question: Are other relationships, such as close friendships or kin ties, sufficiently distinctive that a romantic pair bond cannot compensate for their loss? Ideologically, that question is out of bounds. It seems to us that the fundamental theoretical issue is the delineation of the qualities that set adult attachment relationships apart from other adult relationships. Not all marital relationships will qualify, and some friendships and relationships with relatives will. Once a relationship does become an attachment relationship—regardless of whether it is a sexual partnership—it may well be irreplaceable.
One more observation from Hazan and Zeifman (1999) about sexual pair bonds is also worth noting:
“As for the protective aspects of this kind of companionship, adults too need someone to look out for them and keep track of them–someone to initiate a search if they fail to show up at the expected time, to care for them when they are sick, dress their wounds, help defend them against external threats, reassure them, and keep them warm at night” (p. 348).
We do not doubt that stable sexual partnerships can, at their best, fulfill such functions, and that doing so is a good thing. What grabbed our attention was the dreamy prose that thoughts of coupling seemed to inspire, and the strange implication that people without a stable sexual relationship are wandering adrift with open wounds and shivering in their sleep. At a minimum, Hazan and Zeifman’s (1999) conclusions need to be re-evaluated. We hope the next review of the issues will have the benefit of studies designed specifically to address the relevant claims.