The most memorable lecture of my college career was the one in which biological anthropologist Melvin Konner asked a room full of impressionable undergraduates what the purpose of love could be. Theatrically, he let a few of us answer. Then he dropped the hammer.

“What if the purpose of love isn’t getting people into relationships, but out of them?” asked Konner, author of more than a half-dozen books on human nature. Think about it, he urged. Love makes us irrational. And what’s more irrational (in a universe in which there are surely more bad possibilities than good ones) than leaving the safety of an existing relationship?

Konner’s question, as a matter both philosophical and moral, isn’t one that science can answer. But science, specifically biological anthropology, can tell us how and why humans tend to act the way they do when they fall in love.

Love 101

The traditional, Western view of romantic love — what philosopher Aaron Ben-Zeév calls the Romantic Ideology— maps to a phenomenon that psychologists call limerence. Whatever you call it, it’s all about the initial bliss and excitement of falling in love. And while expressions of limerence are defined by culture, the underlying process is probably evolutionarily quite old, and exists in many mammals.

Falling in love, or something like it, has been well characterized in monogamous prairie voles, for example. In these animals, a series of clever experiments established that the hormones released when two prairie voles mate — oxytocin and vasopressin — bind to neurons in the part of their brains responsible for reward. These are the same areas activated by drugs of addiction, leading scientists to say that, in effect, drugs hijack systems that evolved to allow us to fall in love.

And while the neurobiology of love may be analogous in humans to the way it works in voles, we don’t understand it nearly as well. There is a limit to what this line of inquiry can tell us about the nature or “purpose” love. For more concrete answers, we have to look to the consequences of love — those things that are easily defined and measured. One of the primary consequences of love, in humans as well as other mammals that trend monogamous, is the “pair bond”. In people, we have great data on getting together and breaking up.

Humans are not monogamous

Monogamy is rare in animals. Only about 3–5% of them practice lifelong monogamy. In humans, it’s only slightly more common. If you look at pre-industrial cultures, which until about 50 years ago meant most of the cultures on earth, 80% practiced some sort of non-monogamy.

There is a spectrum of thinking on the degree to which humans were historically non-monogamous. A popular book called Sex at Dawn argues, essentially, that humans used to live in a halcyon time of polyamorous pre-jealousy that was ruined by the black cloud of patriarchy, agriculture, and monotheism. But as David P. Barash, professor of psychology at the University of Washington contends,the reality is a lot less idyllic. The overwhelming majority of cultures have, in the past, practiced polygyny, a subspecies of polygamy in which only men have more than one partner.

While the Victorian era represents a brief abstention from polygamy, at least in the West, current divorce rates indicate humans are headed right back to historical norms (at least in rich countries) in terms of the number of partners we will have across a lifetime. In the US, experts estimate that between 40% and 50% of all marriages will end in divorce. That’s not atypical for countries with similar levels of per-capita income. And this doesn’t even include the long merry go round of of pre-marriage relationships that has become the norm.

In all of these couplings, re-couplings and de-couplings, we see the footprint of love itself — clearly, something is driving humans to change dance partners with such frequency.

Humans tend to be serially and socially monogamous

Even animals that bond with a single partner for life, like the monogamous prairie vole, will engage in “extra-pair copulations” that may result in offspring. This sort of behavior is common in birds, where a minority of eggs in a clutch laid by a mother in a monogamous relationship will often be fertilized by a male other than her partner. Animals that stick together, but might stray when it comes time to make offspring, are known as “socially monogamous”.

In his lectures, Konner also characterized humans as “serially monogamous,” a nod to the fact that across cultures, humans tend to maintain strict or merely social monogamy with a series of partners.

The miracle here, as some have pointed out, is not that humans are sometimes non-monogamous, but that compared to the rest of the animal world, humans are monogamous at all. Once again, we see evidence of the nature of love in humans: It binds us, if only for a time.

But humans are flexible by nature

Some research suggests the tendency to stray is written into the genes of both women and men. But the incredible diversity of human mating systems — from modern-day sanctioned polygyny to voluntary monogamy — shows that our cultural and psychological flexibility allows us, if not true free will, then at least some agency in how we conduct our relationships.

For evidence of that flexibility, we have only to look at rates of infidelity, which in the US are changing in women but not men. This, in part, is tied to the increasing economic and social power of women. (In both women and men, the link between infidelity and power appears to be confidence.) As humans with higher-order reasoning, and the ability to plan for the future and prioritize things other than our base instincts, we are capable of choosing (or being forced to choose) between priorities other than love.

So what is the purpose of love?

What we don’t know so much about, in the evolutionary sense, is the “purpose” of love. When Konner asked his class what love was for, he meant, as a biological anthropologist, what is its utility in the production of offspring? But as he often reminded us, such questions are inherently unanswerable, since evolution is not a process with the kind of will or intentionality these questions imply. In his scientific world, at least, there was no creator guiding the process of humans’ long narrative across the millennia—there was no “purpose” to any of this evolution, only its results.

But is it unreasonable to suppose that, in the sense most of us understand it, one of the “purposes” of love, not incompatible with the binding together of two people, is to make them crazy enough to ditch their current partners first? Certainly, the overwhelming evidence from our genes and from the history of human societies is that something is driving breakups just as powerfully as that same mechanism, or some related one, drives people to get together in the first place.